X

Recent Posts

TechDay in Vienna: Oracle on Oracle

Following up on the success of the quite successful TechDay in Vienna: Solaris event we scheduled the next TechDay with the focus "Oracle Software advantages on Oracle Hardware".  *drumroll*Ladies and Gentlemen, allow me to announce the  TechDay in Vienna:Oracle on Oracle event, on the afternoon of January 30!  Location: 25hours Hotel, Vienna Registration link: here! Short summary:  This event follows up on the success of the Solaris TechDay Vienna, this time with emphasis on Datacenter Technologies. We will focus on the practical implementations of the integration between the Database, Virtualization-, Server-, Network- and Storagetechnologies, discussing the technical background, providing detailed examples as well as live demonstration of the mentioned technologies.  We invite Database Administrators, Storage- and System Engineers, Platform Architects and Operation Leaders to join us in this free, english-speaking, half-day long international technical event, exchange experiences with partners and other customers, learn about current technologies to enrich your business department's experience running critical databases and applications with efficience, performance and ease. Agenda Draft: 12:45 Registration 13:15 Opening and Welcome: The Story Roman Brandl, Oracle 13:15 Oracle Systems: Server-, Network- and Storagetechnologies for the DB stack  Karoly Vegh, Oracle 14:10 Database operation options on Oracle Systems Daniel Hafner, DBConcepts 14:35 Coffee break 14:50 Managing Solaris Virtualization with JomaSoft's VDCF  Marcel Hofstetter, JomaSoft 15:15 Fast Database and Application instance deployment with EM12c and ZFS snaps  Agnes Kovacs, Oracle 15:40 Customer Case Study: How Verbund is running a HighAvailable SPARC/Solaris Virtualization platform for Oracle DataBases  René Biedermann, Verbund 16:05 Sessions Wrap-Up Karoly Vegh, Oracle 16:15 Coffee break/snacks 16:35  Parallell Breakout Sessions: LiveDemo: Virtualisation on SPARC with Logical Domains LiveDemo: Combine Zones and LDoms with VDCF LiveDemo: Enterprise Manager 12c and Database-as-a-Service with Storage Integration LiveDemo: Oracle DB 12c exclusive features on SPARC LiveDemo: Managing the ZFS Storage Appliance: Overview, Administration, Performance Analytics  Stefan Hinker, OracleMarcel Hofstetter, JomaSoftAgnes Kovacs, Oracle Daniel Hafner, DBConcepts Lubomir Vano, Oracle  17:45 Open End: Networking by snacks and drinks Should you have questions, do not hesitate to mail or tweet me with the hashtag #TechDayinVienna. Looking forward to seeing you there! 

Following up on the success of the quite successful TechDay in Vienna: Solaris event we scheduled the next TechDay with the focus "Oracle Software advantages on Oracle Hardware".  *drumroll*Ladies...

Infiniband: a highperformance network fabric - Part I

Introduction:At the OpenWorld this year I managed to chat with interesting people again - one of them answering InfiniBand deepdive questions with ease by coffee turned out to be one of Oracle's IB engineers, Ted Kim, who actually actively participates in the InfiniBand Trade Association and integrates Oracle solutions with this highspeed network. This is why I love attending OOW. He granted me an hour of his time to talk about IB. This post is mostly based on that tech interview.Start of the actual post: Traditionally datatransfer between servers and storage elements happens in networks with up to 10 gigabit/seconds or in SANs with up to 8 gbps fiberchannel connections. Happens. Well, data rather trickles through.But nowadays data amounts grow well over the TeraByte order of magnitude, and multisocket/multicore/multithread Servers hunger data that these transfer technologies just can't deliver fast enough, causing all CPUs of this world do one thing at the same speed - waiting for data. And once again, I/O is the bottleneck in computing. FC and Ethernet can't keep up. We have half-TB SSDs, dozens of TB RAM to store data to be modified in, but can't transfer it. Can't backup fast enough, can't replicate fast enough, can't synchronize fast enough, can't load fast enough. The bad news is, everyone is used to this, like back in the '80s everyone was used to start compile jobs and go for a coffee. Or on vacation. The good news is, there's an alternative. Not so-called "bleeding-edge" 8gbps, but (as of now) 56. Not layers of overhead, but low latency. And it is available now. It has been for a while, actually. Welcome to the world of Infiniband. Short history:Infiniband was born as a result of joint efforts of HPAQ, IBM, Intel, Sun and Microsoft. They planned to implement a next-generation I/O fabric, in the 90s. In the 2000s Infiniband (from now on: IB) was quite popular in the high-performance computing field, powering most of the top500 supercomputers. Then in the middle of the decade, Oracle realized its potential and used it as an interconnect backbone for the first Database Machine, the first Exadata. Since then, IB has been booming, Oracle utilizes and supports it in a large set of its HW products, it is the backbone of the famous Engineered Systems: Exadata, SPARC SuperCluster, Exalogic, OVCA and even the new DB backup/recovery box. You can also use it to make servers talk highspeed IP to eachother, or to a ZFS Storage Appliance. Following Oracle's lead, even IBM has jumped the wagon, and leverages IB in its PureFlex systems, their first InfiniBand Machines.IB Structural Overview: If you want to use IB in your servers, the first thing you will need is PCI cards, in IB terms Host Channel Adapters, or HCAs. Just like NICs for Ethernet, or HBAs for FC. In these you plug an IB cable, going to an IB switch providing connection to other IB HCAs. Of course you're going to need drivers for those in your OS. Yes, these are long-available for Solaris and Linux. Now, what protocols can you talk over IB? There's a range of choices. See, IB isn't accepting packet loss like Ethernet does, and hence doesn't need to rely on TCP/IP as a workaround for resends. That is, you still can run IP over IB (IPoIB), and that is used in various cases for control functionality, but the datatransfer can run over more efficient protocols all mapped directly on native IB. About PCI connectivity: IB cards, as you see are fast. They bring low latency, which is just as important as their bandwidth. Current IB cards run at 56 gbit/s. That is slightly more than double of the capacity of a PCI Gen2 slot (of ~25 gbit/s). And IB cards are equipped usually with two ports - that is, altogether you'd need 112 gbit/s PCI slots, to be able to utilize FDR IB cards in an active-active fashion. PCI Gen3 slots provide you with around ~50gbps. This is why the most IB cards are configured in an active-standby way if both ports are used. Once again the PCI slot is the bottleneck. Anyway, the new Oracle servers are equipped with Gen3 PCI slots, an the new IB HCAs support those too. Oracle utilizes the QDR HCAs, running at 40gbp/s brutto, which translates to a 32gbp/s net traffic due to the 10:8 signal-to-data information ratio.  This means for some peak bandwidth cases, active-active must be used with QDR to get all the slot bandwidth.Consolidation techniques: Technology never stops to evolve. Mellanox is working on the 100 gbps (EDR) version already, which will be optical, since signal technology doesn't allow EDR to be copper. Also, I hear you say "100gbps? I will never use/need that much". Are you sure? Have you considered consolidation scenarios, where (for example with Oracle Virtual Network) you could consolidate your platform to a high densitiy virtualized solution providing many virtual 10gbps interfaces through that 100gbps? Technology never stops to evolve. I still remember when a 10mbps network was impressively fast. Back in those days, 16MB of RAM was a lot. Now we usually run servers with around 100.000 times more RAM. If network infrastrucure speends could grow as fast as main memory capacities, we'd have a different landscape now :) You can utilize SRIOV as well for consolidation. That is, if you run LDoms (aka Oracle VM Server for SPARC) you do not have to add physical IB cards to all your guest LDoms, and you do not need to run VIO devices through the hypervisor either (avoiding overhead). You can enable SRIOV on those IB cards, which practically virtualizes the PCI bus, and you can dedicate Physical- and Virtual Functions of the virtualized HCAs as native, physical HW devices to your guests. See Raghuram's excellent post explaining SRIOV. SRIOV for IB is supported since LDoms 3.1.  This post is getting lengthier, so I will rename it to Part I, and continue it in a second post. 

Introduction:At the OpenWorld this year I managed to chat with interesting people again - one of them answering InfiniBand deepdive questions with ease by coffee turned out to be one of Oracle's IB...

So, what makes Solaris Zones so cool?

How do you virtualize? Do you emulate virtual machines? Do you partition your servers' hardware? Or do you run a container technology? This post is about the third option, a container technology built right into Solaris: Solaris Zones. They are pretty awesome, especially on Solaris 11 - they're like vacation: once you go Zones, you won't want to leave them :) But what exactly makes Zones so cool? There are a number of reasons, allow me to list my favourite top 10: Pro primo: Solaris Zones are performance overheadless. That is, you will not lose any CPU capacity due to virtualization. The reason for this container feature is that there is no additional emulation/virtualization layer between the bare-metal serverinstallation and the Zone instances - because all the kernel-level services (resource mgmt, I/O, scheduling, etc.) are provided by the kernel running as part of the global zone (=="bare metal" installation) and the Zones themselves are userspace containers only! Pro secundo: Resource Management. Resource management has been around since (IIRC) Solaris 9 with projects, but it turned out to be immensely useful with Zones. You can of course create static CPU pools, bind some zones to each, but my favourite method was using the FairShareScheduler (FSS), that allows definining a guaranteed minimum amount of CPU time portion per zone, but allowing the zone to consume way more as long as the CPU pool isn't 100% utilised. Also, see Stefan's post about memory DR for zones. Pro tertio: the Golden Image cloning: It makes a lot of sense putting the zones to ZFS due to ZFS's natural cloning capabilities. What I usually demo to customers is cloning zones. That is, I have a prepared zone that serves as a golden image (the source of cloning) and instead of installing the next zone from the IPS repository, I clone the existing one with zoneadm's built-in cloning capabilities. In a matter of seconds. And that's differential within the zpool, that is, it uses hardly any space on the disks at all. Pro quatro: Zone independence: Zones are (especially since Solaris 11) pretty much independent, they [can] have their own: users (including a separate root user per zone)  configurations  process space that isn't visible to other nonglobal zones filesystems and mounts  software packages  SMF services  exclusive IP stack  bootenvironments(!) namespace That makes zones independent to a large extent without having to implement an own kernel layer on top of the kernel that is already running in the global zone.  Pro quinto: Branded Zones. Yes, you can run a complete Solaris 10 zone on top of Solaris 11. Within the zone you will have complete Solaris 10 userspace archived from an existing system that was either a Solaris 10 global zone or a non-global one. Just like that. Pro sexto: Cluster integration: Solaris Zones can be moved around between nodes of a Solaris Cluster as "moving zones" with the built-in clusteragent delivered by Solaris Cluster (former SunCluster). But with Solaris Cluster you also can create ZoneClusters, that is, clusterinstances on the zone level. A zonecluster includes a non-global-zone created and managed by the Global Cluster on each clusternode, while the clustercommands are available within those zonecluster-nodes. That is, you can have several zoneclusters on top of the "bare metal" Solaris Cluster installations, having separate, independent, multitenant, virtual clusterenvironments. For more details see this post: "Zones? Clusters? Clustering zones? Zoneclusters?".You can even have Solaris-10-branded-zoneclusters! Pro septimo: Immutable Zones: there are different predefined levels of setting parts of a zone read-only from the inside. The filesystems are read-only from the nonglobal zone, but writeable from the global zone - hence providing an additional line of defense for the global zone's administrator in case the zones are managed by other zoneroot-users (see the documentation for details, or see Darren's blogpost about an encrypted immutable zone on iSCSI).  Pro octavo: Exclusive IP stack with VNICs: on Solaris 11 you can have your cake and eat it too: All your zones are by default exclusive-IP-stack zones, that is, they own a separate Level3 stack, that only belongs to that zone, eliminating shared network spaces between zones. And you don't even need to dedicate HW NICs to the zones for that, since in Solaris 11 thanks to project Crossbow you can create virtual NICs (vnics) and dedicate those L2 elements to the zones. The zone can even manage it's own network configuration - and yes, there is network spoofing protection built right into zones and vnics. You can even run an NFS server in a zone now! For details, see Jeff Victor's excellent Solaris-10-Zones-compared-to-Solaris-11-Zones post.Pro nono: zonestat: Zonestat is a tool that provides a good overview about what resources and to what extent your zones are utilizing, in a readable/parseable way. By default it shows CPU, memory and network utilization, just try "zonestat 5 3" on a system with busy zones before you'd dive into additional parameters.  Pro decimo: Per-zone fstype statistics: as discussed in the post "What's new in Solaris 11.1?" zones got quite an update in Solaris 11.1 (and development does not stop!), but one of the main features was to be able to print filesystem statistics with "fsstat" on a per-zone basis. Of course there are more than 10 mentionworthy features we haven't mentioned, these were my top 10. We didn't talk about Zones on Shared Storage, Parallell updates, etc. What are your favourite features? As usual, should you have questions: do not hesitate to ask in the comments. -- charlie 

How do you virtualize? Do you emulate virtual machines? Do you partition your servers' hardware? Or do you run a container technology? This post is about the third option, a container technology...

A SPARC CPU Overview: SPARC64-VII+ (M3), T3, T4, T5, M5, M10 - how do they compare?

Do you know how many SPARC CPU types does Oracle deliver currently? Not one, not two, but five. The SPARC64-VII+, the T4, T5, M5 and M10. It isn't easy to keep track of these, hence we created an overview table to help you in that - here you are:  CPU Type  SPARC64-VII+ (M3)  T3 T4 T5 M5  M10/SPARC64-X Servers M3000, M4000,M5000, M8000,M9000, M9000-64 T3-1, T3-2T3-4  T4-1B, T4-1, T4-2T4-4 T5-1B, T5-2, T5-4T5-8 M5-32 M10-1, M10-4, M10-4S CPU Core Codename Jupiter++ S2 S3 S3 S3 Athena CPU Clockfrequency 2.66-3 GHz 1.65 GHz 2.85-3GHz 3.6 GHZ 3.6 GHZ  2.8-3 GHz Number of Cores per socket 4 16  8  16 6 16 Number of Threads per core 2 8  8  8 8 2 Minimum amount of RAM,number of sockets, and threads per Server  8GB1 socket8 threads 8GB1 socket128 threads  16GB1 socket64 threads 128GB1 socket128 threads 1TB8 sockets384 threads 32GB1 socket4 threads (2 cores activated from 16) Maximum amount of RAM, number of sockets and threads per Server 4TB64 sockets512 threads  128 GB4 sockets512 threads 2TB4 sockets256 threads  4TB8 sockets1024 threads 32 TB32 sockets1536 threads 32TB64 sockets2048 threads L3 Cache no L3$(12MB L2$ though!) no L3$ 4MB 8MB 48MB no L3$(24MB shared L2$ though!) Virtualization supported Dynamic Domains(and Zones) LDoms(and Zones) LDoms(and Zones)  LDoms(and Zones)  Hard Domains, LDoms(and Zones)  Physical Partitions (on 4S), LDoms(and Zones)  CPU architecture (ISA) sun4u sun4v sun4v sun4v sun4v sun4v  OS supported Solaris 10, 11 Solaris 10, 11 Solaris 10, 11 Solaris 10, 11 Control Domain: S11 only. Guest Domains: Solaris 10, 11 Solaris 10, 11 noteworthy features electronically separated domains 512 threads in 2010! First S3 based, single thread AND throughput 1024 threads in 8 RU Mainframe class, Mission critical Software on Chip Allow me to add some thought-triggering facts too: All the 6 of these CPUs have been released after Oracle taking over Sun. (Oracle also has released Solaris 10 Update 10, Solaris 10 u11, Solaris 11 and Solaris 11.1 since then!)  CPU cache is important: reading from RAM takes orders of magnitudes longer than from on-chip cache.  It takes 5 years from CPU design start to deliver. Here's a quick thought-experiment:  IF the T5 was already deep in the design phase before the acquisition in 2009, AND Oracle raised the investment into SPARC development THEN what can you assume about the next 1-3-5 years in terms of SPARC products?  should you have questions, as usual, do not hesitate to ask in the comments! 

Do you know how many SPARC CPU types does Oracle deliver currently? Not one, not two, but five. The SPARC64-VII+, the T4, T5, M5 and M10. It isn't easy to keep track of these, hence we created...

Achievement Unlocked: Successful Solaris TechDay in Vienna. Your Task: Feedback Please!

Update: For the event's presentation slides please scroll down to the bottom of the post. The Solaris TechDay 2013 in Vienna was outstandingly successful.Some characteristics: Event sized for: maximum 80 attendants. Registrations: Over 110. Number of participants who really showed up: 98. We had customers, partners, Oracle colleagues visiting from Austria, Hungary, Croatia, Romania, Slovenia! Pretty impressive :) We started with the technical presentations, we heard a keynote explaining that Oracle runs best on Solaris from Solaris Product Manager Joost Pronk, then saw an update about what's new in Solaris 11.1 by Joerg Moellenkamp, learned about Arrow ECS' Solaris Training Center, heard about real-life implementation scenarios with SAM/QFS and Solaris Cluster from Rudi Rotheneder, and discussed how Solaris Zones integrate with Solaris Cluster by Karoly Vegh.  After the presentations we distributed our technical attention between four Hands-on-Labs (Zones and ZFS) and LiveDemos (LDoms by Rudi and Solaris Performance Visualization by Manuel Zach).  After those several discussions spun of by coffee, snacks and beer, many already proudly wearing their freshly acquired Solaris Day TShirts.  And now it is your turn, we need your feedback! Please let us know what you thought of the event. We are looking for both positive feedback about what worked well for you on the TechDay, and also constructive criticism. We would like to hear what was new for you, what you would like to hear about a possible future event.  Feel free to add your comments below, send me a mail directly, or tweet me, about presentation content, event execution, wishes, techtopic requests, how the HoLs and the Demos worked out for you, or anything you'd like to say about the event.  Feedback is pretty important. Let us know how to support the Solaris community further. This is the only way to keep up events like this and improve them further. As soon as we have 10 feedback posts I'll release the presentation slides! :)  Thank you in advance! -- charlie  UPDATE: Here come the slides:   Title   Slides  Presenter Oracle Solaris - Features, Oracle on Oracle Optimizations, and Futures Download Joost Pronk, Solaris Product Manager What's new in Solaris 11.1? Download Jörg Möllenkamp, Senior Account Architect Partner Session I: Arrow ECS: Empowering the Solaris Community Download Business Development Manager Software  Partner Session II: cons4you: Real World Solaris Experiences: Advanced Technologies Download Ing. Rudolf Rotheneder CEOcons4you GmbH. Solaris Cluster: High Availability options integrating with Solaris Virtualization Download Karoly Vegh, Principal Systems Consultant Demo: Visualizing Solaris performance the easy way Link Manuel Zach, System Engineer

Update: For the event's presentation slides please scroll down to the bottom of the post. The Solaris TechDay 2013 in Vienna was outstandingly successful. Some characteristics: Event sized for:...

Solaris Technical Day on the 23rd May in Vienna

Update: the registration link is live! What is the best way to directly exchange information with your fellow engineers running UNIX platforms or applications thereupon? In our opinion it is a face-to-face chat by a conference-coffee or by an after-conference beer. To enable this kind of less formal technical information exchange we set up a technical event, the Solaris Day 2013 in Vienna.  We invite customers, partners, SystemEngineers, DBAs, ApplicationEngineers, and anyone else interested in Solaris technologies to join us on the 23rd May, in Vienna in the Urania to review past, present and future Solaris technologies, discuss real-world customer usecases, learn about the ways Solaris features are utilized in different scenarios.  What do you need to know? I.    The event is on May 23, in Vienna, in the Urania.II.   The presentations will be held in english, for we have invited and expect an international audienceIII.  The registration link is live. Also, save the date into your calendar (do it now! 23rd May, 12:45-19:00) IV.  To participate in the technical Hands-on-Labs bring along your laptop with a 64-bit OS and an up to date VirtualBox installation (can be done withing minutes).V.   This is a free event, only registration is necessary to reserve resources.VI.   We are looking forward to seeing you!   Here's the agenda to raise your interest:  kvegh@s11u2:~$ cat projects/Solaris_Day/agenda.txt12:45: Registration 13:15: Opening and Welcome  13:30: Oracle Solaris - Features, Oracle on Oracle Optimizations, and Futures - Joost Pronk, Solaris Product Manager 14:10: What's new in Solaris 11.1? - Jörg Möllenkamp, Senior Account Architect  14:50: CoffeeBreak 15:10: Partner Session I: Arrow ECS: Empowering the Solaris Community 15:30: Partner Session II: cons4you: Real World Solaris Experiences: Advanced Technologies - Rudolf Rotheneder, CEO cons4you 16:20: Solaris Cluster: High Availability options integrating with Solaris Virtualization - Karoly Vegh Principal Systems Consultant 17:00: Parallell Tracks:         - Hands-on-Lab: Zones - lead by Karoly Vegh         - Hands-on-Lab: ZFS - lead by Jörg Möllenkamp         - Demo: LDoms - presented by Rudolf Rotheneder        - Demo: Visualizing Solaris performance the easy way - delivered by Manuel Zach        - Networking, informal information exchange kvegh@s11u2:~$ Update 2: - As you can see we have managed to invite/fly in quite a list of distinguished guest presenters:  Joost Pronk is a Solaris Product Manager. You know, one those guys actually forming Solaris. Jörg Möllenkamp is the well-known author of the c0t0d0s0.org Solaris techblog, and the writer of the Less Known Solaris Features doc Rudolf Rotheneder is the head engineer of the austrian partner company cons4you, maintaining many Solaris Cluster and SAM/QFS platforms in the region Manuel Zach is a highly skilled System Engineer, techblogger, established Solaris betarelease-tester, focusing on advanced technologies like cluster infrastructures, DTrace or performance monitoring/tuning. Should you have any questions, to not hesitate to ask! In the comments, on Twitter (use the hashtag #SolDayVienna), or per E-Mail. -- charlie 

Update: the registration link is live!What is the best way to directly exchange information with your fellow engineers running UNIX platforms or applications thereupon? In our opinion it is a...

What is the best platform to run your Oracle Database on?

As a Systems Consultant I am often faced with the following question: What does Oracle recommend as the best platform to run the Oracle DB in large enterprise environments on? Is the Exadata DB Machine the recommended platform? Are SPARC/Solaris servers the way to go? Or should customers consolidate on SPARC SuperCluster setups? ...at this point customers look me deep into the eyes, and ask the critical question: What does Oracle actually recommend?The answer usually perplexes them, for I say every time: "It depends."  And that's the best answer one can give - there is no one-size-fits-all solution. Different customers follow different goals, have different requirements, platforms, expectations, experiences. I prefer to recommend a solution that fits the most their platform, their plans and projects. To answer the question about the optimal platform for DB Operations in a specific scenario there are a number of aspects to take into account, allow me to provide decision support. I will talk about first the Exadata, then SPARC/Solaris servers and last but not least about the SPARC SuperCluster. I. When shall customers go with an Exadata solution? With the Exadata Database Machine, the enterprise DB solution customers pursuing the following goals benefit the most:  A worldwide Standardized DB platform Time to market extremely reduced due to fast deployment Hardware and Sofrware preintegrated and verified into a single platform, avoiding platform development and testing expenditure Maintenance efforts reduced due to coordinated, bundled and verified software releases Option to grow fast in terms of both capacity and computing Option to choose between both Linux and Oracle Solaris Extreme performance thanks to the built-in database-aware storage solution Removing burden from the network and the block-storage II. What customer goals and requirements are addressed well with SPARC/Solaris servers?  There are several other criteria to consider where SPARC/Solaris benefits weigh a lot in platform-choosing decisions. Customers choosing to deploy Solaris servers for Database platform aim to fulfill the following requirements and enjoy the following advantages: Solaris 11 provides unique features specifically for the Oracle Database that are available on Solaris only Running any other non-DB applications, like Java, Webservices, etc. Multitenancy provided by built-in Solaris and SPARC virtualization Oracle-owned SPARC IP being the only platform where Oracle implements Software-in-Silicon features accelerating applications to hardware-speed Possibility to use software-license boundaries via virtualization enabling customers to start small and grow their capacity on demand Solaris and the Oracle Database being co-developed and harmonized for eachother  Multithreaded applications benefiting from the massively multithreaded SPARC architecture Using existing storage and network infrastructure For detailed technical recommendation on building Small/Medium/Large/extraLarge platforms see the Technical Whitepaper of the Oracle Optimized Solution for Oracle Database.  In the latter situation customers have the option to start with standalone SPARC servers that can grow and scale up to the level of an engineered system - a great example is the SPARC SuperCluster, that utilizes exactly these features for an upgrade path for SPARC customers. Which is our next topic anyway: III. In what scenarios should customers choose a SPARC SuperCluster?  Actually a SPARC SuperCluster unites the advantages of an Exadata Database machine and the SPARC/Solaris standalone servers. It provides extreme high performance, the same level of integration, the DataBase Storage Cells, the same standardization like an Exadata setup, and utilizes the virtualization features of SPARC/Solaris, runs Java or any other applications as well, does the heavy multithreading and provides strict multitenancy.  It is the true heavylifting enterprise champion for different applications that incorporate all the benefits listed above. It is even the standard building-block for many Optimized Solutions. For detailed technical recommendation on building Oracle Optimized Solution for Enterprise Database Cloud see the Technical Whitepaper.  In summary we can say that the choosing best platform running your Oracle Database always depends on the requirements and goals of a company. I sincerely hope to have provided a short overview of the benefits to support your decision process. 

As a Systems Consultant I am often faced with the following question: What does Oracle recommend as the best platform to run the Oracle DB in large enterprise environments on? Is the Exadata DB...

The SPARC T5 Servers have landed

At last, the SPARC T5 servers are announced! They are worthy heirs of the SPARC T-series, and even more powerful than their successful predecessors the SPARC T4.  Before we'd go into the technical details, here is a very important message:  Having announced the T5 servers doesn't make the T4s go away. It is not a platform replacing technology, but a platform extending one! Oracle is going to offer SPARC T4 and T5 servers side-by-side! Here's a short overview of the complete SPARC serverportfolio:  Having said that, let's dive headfirst into the details to soothe the techcrave:  What's new with the SPARC T5 servers? The more obvious updates: One CPU has 16 cores  The cores run at 3.6GHz Servers with up to 8 sockets  that is, up to 1024 parallell threads in a box!! 8 sockets wiht glueless interconnect, mind you, no interim hops between the sockets!   up to 4TB memory per server More detailed information: Cache doubled to from the T4's 4MB to 8MB L3 cache  PCIe Slots upgraded to PCIe 3.0 doubling throughput of I/O CPU Modules are hot-pluggable on the T5-4 and T5-8 servers  Crypto performance is more than double to that of a T4  The T5 is using the 28nm technology (T4 was built with 40nm)  Advanced Power Management  T4 had ExpressModule PCIe, T5 has standard lowprofile PCIe Slots We sticked to the many T-Server goodies though:  OVM for SPARC (aka Logical Domains, the possibility to partition the T-SPARC servers to several virtual machines running completely independent Solaris installations/versions)  HW Cryptoengine (encrypt your DataBase data, ZFS filesystems, SSL, support your iSCSI traffic with HW-implemented algorithms)  Solaris 10 and Solaris 11 support  The CPU Cores are the S3 cores we learned to appreciate in the T4s OnChip PCIe controllers  Out-of-Order, dual-issue instructions Oracle has released 1-, 2-, 4- and 8 Socket servers, the T5-1B (blade form factor), the T5-2, the T5-4 and the T5-8. These have been delivered as per the SPARC Roadmap, which Oracle has committed to not to deliver later than defined.  These expand the range of the SPARC T-servers, starting with 4-core Netra SPARC T4-1 all the way up to 128 cores SPARC T5-8. These are the servers you are looking for, may the force be with us to feed these massively parallell enterprise beasts. Questions? Do ask! Like: Q: You didn't mention yet the M5...? A: I didn't dare to put too much awesome into one single post. Allow me to come back and report in another one :) Yes, it has up to 32 Sockets, 1536 parallelly running threads, and up to 32 TB RAM. Yes it runs Hard Partitions, LDoms, Solaris 10 and 11 :) 

At last, the SPARC T5 servers are announced! They are worthy heirs of the SPARC T-series, and even more powerful than their successful predecessors the SPARC T4.  Before we'd go into the technical...

Best Practices to build SAP, Tiered Storage, IaaS cloud, E-Business Suite or Siebel HW platforms? Optimized Solutions.

Uncountable IT Operations departments build platforms for SAP. Thousands of Administrators design tiered storage solutions. Too many engineers spend their time integrating hardware setups for E-business Suite. A huge number of Architects are involved in creating private IaaS cloud services. An army of SystemManagers work on Siebel server landscapes. Having said that you probably expect me to state that they are all wrong. I won't - generally, systemengineers build platforms of quite high quality. But I do state that most of these projects are either hopelessly undersized with headcount and generally take way longer to implement than they were supposed to. I can't give you a solution (besides probably recommending Advanced Customer Services) to raise the headcount. Nearly every company struggles in that area. However I can tell you that most projects suffer heavy delays in the same phase of the projects: From looking through the vast array of best of breed components and then testing, validating and integrating the various building blocks of the future platform. Exactly this effort can be minimized. Oracle provides a number of freely available whitepapers, describing HW architecture best practices combining Sun servers, storage and networking components, their configurations, the virtualization and the OS setup on them, the application structures for different Applications. These include SAP, Oracle HCM, Weblogic, the Oracle database or providing guidelines to build tiered storage platforms or Infrastructure-as-a-Service cloud environments and a number of other architectures.  These are called Oracle Optimized SolutionsThese solutions are already tested, validated, integrated and recommended by Oracle. Following them companies can avoid the initial longsome evaluation and integration process of different third party elements.  So, exactly what type of solutions/platforms are described in these Optimized Solutions? Here's an overview:   Optimized Solution  Remarks Oracle Optimized Solution for Enterprise Cloud Infrastructure There is a SPARC and an x86 version of this paper  Building blocks: SPARC T4 or x86 servers, ZFS SA with 10gbps Ethernet, Oracle VM for SPARC (and x86, respectively), EM Ops Center for management The whitepaper covers: Virtualization setup, failover/livemigration concepts, Ops Center installation, Monitoring.  Oracle Optimized Solution for Tiered Storage Infrastructure Put Disk, NFS and Tape capacities into a single filesystem, export it via NFS or CIFS to clients The PDF covers: Small, Medium, and Large setups Oracle Optimized Solution for SAP Building blocks: 2x SPARC SuperCluster, outer connection: 10gbps Ethernet The whitepaper covers: Virtualization Setup, Solaris Cluster setup, InfiniBand connections, SAP installation, DB and RAC installation.  Oracle Optimized Solution for Oracle E-Business Suite Building blocks: T4-2, T4-4 servers or SPARC SuperCluster depending on the expected load The whitepaper covers: Virtualization setup, HA setup, Content staging between test and production instances, small-, medium- and large configurations.   Oracle Optimized Solution for Oracle WebLogic Server Building blocks: T4 and x86 servers, 10gbps Ethernet, Solaris 10 containers, WebLogic clustering  The whitepaper covers: Virtualization setup, network tuning, network traffic encryption offload, S/M/L/XL setups General Page enlisting all the Oracle Optimized Solutions There are many more, like for Peoplesoft HCM, Webcenter Portal, Oracle Unified Directory, etc. Please don't make me list them all here :)  To sum it up: if you plan to build an IaaS cloud, spare costs with real Storage Tiering, implement an SAP platform, a Siebel CRM setup, a WebLogic serverfarm, or many of the other common softwareproducts used worldwide, use these preintegrated and recommended architectures to cut the projects efforts needed impressively. 

Uncountable IT Operations departments build platforms for SAP. Thousands of Administrators design tiered storage solutions. Too many engineers spend their time integrating hardware setups for...

How to build a private IaaS cloud platform? - Part IV, Shared Storage

This is Part IV of the series "How to build a private IaaS cloud platform?", elaborating about the platform requirements to face when building a multipurpose Infrastructure as a Service platform. In this post we are looking into the requirements the chosen shared storage has to fulfill. See the Main Summary Post with links to the other sections (requirements analysis for serverplatform, OS, virtualization, etc) here.  Part IV: Shared Storage So, what are the most important requirements a shared storage has to fulfill to be chosen as an IaaS cloud building block? Requirements list: Snapshots/Cloning: Running hundreds of virtual machines in a shared environment quickly triggers the thought of how convenient it is to snapshot a state of an installed VM on the storage and keep its state where the admins can roll back to any time. Also, having snapshots allows one to clone VMs, hence quickly roll out or duplicate new environments. Having differential clones of a 'golden copy' allows one to spare a significant amount of storage The reason we see snapshotting and cloning as hard requirements in IaaS platforms is that they enable administration, speed up deployment and provide a safety net in a quickly changing environment.  Deduplication: Another way of keeping storage utilization high is having deduplication on the storage level among the virtual machines installation images. These VMs - most of them sharing probably large amounts of the same binaries - have lots of storage blocks used to store duplicate records. If the storage can deduplicate these, one can practically oversubscribe the storage utilization.  Replication: Today's shared storage products all offer availability features. Nonetheless, complete outages happen, wether taking loss of power, fire, quakes, vogon spacefleets into account, customers prefer to have disaster recovery sites, where they replicate their productive data - including their current VM snapshots - to. Replication provides availability of VMs in disastrous events.  Storage Tiering: Again, costs. Most customers prefer not to have to fill up their storage arrays with solid state disks (SSDs) for those - while delivering staggering performance - are usually bound with higher costs. They prefer to mix SSDs, high throughput and high capacity disks and in some cases Tape in their shared storage implementation and choose to rely on the intelligence of their setup to shuffle data around on demand. That is, they expect storage autotiering within the box for costefficience. Storage Tiering saves costs, hence it is a usual requirement for an IaaS storage backend.  Application Support: The ever growing, unsatisfiable demand for I/O performance of today's applications cannot simply be provided by raw amounts of read/written blocks. The storage shall be closer integrated with the applications supporting their features, enabling administration and performance.  Network Fabric Bandwidth: Often not the disk backend, not the computing frontend, but the connection between them, the network fabric is the performance bottleneck in a platform. The bandwidth of this shall be high, with as low latency as possible.  Observability: Complex storage setups run several workloads of different characteristics parellelly. Analyzing storage-behaviour in the storm of I/O requests can quickly get pretty challenging, although manageability, observability is one of the main factors of Storageadministration in terms of efficiense.  Alright, we got this far. We have some important requirements to fulfill. Let me introduce you to the ZFS Storage Appliance in case you were not familiar with it - and also, allow me to show you how it fulfills these requirements:  Sun ZFS Storage Appliance: ZFS Hybrid Storage Pools Database integration FC/10g/Infiniband DTrace  Deduplication Clustered Heads How the ZFS Storage Appliance fulfills these requirements is: Snapshots/Clones: The ZFS SA, as you probably have guessed, leverages the features of ZFS. In the ZFS world clones and snapshots are self-evident. That is, snapshotting VMs (for later rollback of for cloning) is supported as a matter of course on the ZFSSA.  Deduplication: The ZFS SA, as you probably have guessed, leverages the features of ZFS. In the ZFS world deduplication is self evident. That is, running hundreds of VMs that are built on more or less the same binaries allows administrators to save storage capacity by deduplicating the storage below them.  Replication: The ZFS SA, as you probably have guessed, leverages the features of ZFS. In the ZFS world replication is self evident. Systemadministrators will talk about ZFS's send/receive capabilities, that allows them to replicate data from one ZFS pool to another one. The replication framework of the ZFSSA is based on this capability too, enabling customers to replicate important VMs to a DisasterRecovery site.  Storage Tiering: The ZFS SA, as you probably... you know the drill. ZFS has the Hybrid Storage Pool capability, which allows it to utilize DRAM, SSDs and HDDs in the same pool, caching both read and write data, selftuning the storage performance.  Application Support: Using the application example databases:  The Oracle database can connect to NAS storages with Direct NFS, shortcutting the I/O path between database and storage, improving performance  The ZFS SA is certified for HCC (Hybrid Columnar Compression), a very efficient DB compression method to spare data storage.  OISP: The Oracle Intelligent Storage Protocol that is going to tag I/O operations with additional information instructing the Storage Appliance to handle diverse I/Os differently Network Fabric Bandwidth: Next to the traditional FiberChannel and Ethernet connections the ZFS SA supports connections over InfiniBand too, that as of now provides very low latency with 40gbps raw throughput per cable. Eliminate the network bottleneck. Yes of course you can run IP over IB. Also, with Oracle Virtual Network (former XSigo) you can virtualize FC connections over IB too.  Observability: The ZFS SA provides an we-based management interface that contains DTrace Analytics - a live storage-operations/performance analytics framework generating graphs based on filtered events on-the-fly. Have you ever tried to get statistics out of your storage broken down to IP/user/protocol/disk basis? With the ZFS SA's DTrace Analytics it's easy. Reduce maintenance efforts by enabling your administrators with storage observability!  SUMMARY The ZFS Storage Appliance fulfills the most common requirements by IaaS platforms by providing VM-management features like deduplication, cloning and replication, providing performance enhancement capabilities by Hybrid Storage Pools and InfiniBand support and keeping maintenance efforts by the live observability framework DTrace Analytics.  In the previous post we had a look into the requirements for Operating Systems for IaaS platforms, and how Solaris 11 fulfills those. In the next post we will look into the requirements for a central management platform for IaaS platforms.  If questions arise, do not hesitate to comment.   -- charlie 

This is Part IV of the series "How to build a private IaaS cloud platform?", elaborating about the platform requirements to face when building a multipurpose Infrastructure as a Service platform.In...

How to build a private IaaS cloud platform? - Part III, Operating System

This is Part III of the series "How to build a private IaaS cloud platform?", elaborating about the platform requirements to face when building a multipurpose Infrastructure as a Service platform. In this post we are looking into the requirements the chosen OS has to fulfill. See the Main Summary Post with links to the other sections (requirements analysis for CPU/Server features, shared storage, virtualization, etc) here.  Part III: Operating System So, what are these common requirements an OS has to fulfill to be chosen as an IaaS cloud building block? Requirements list: Application Separation: Every customer wants to have their applications in separated, self-contained environments, but noone wants to have separate VM for each and every of their applications, fragmenting their resources. To fulfill this balancing act, the OS running in the VM shall provide a facility to logically segregate the environments for different applications.  Safe Updates: Updating the OS in the VMs has to happen in a fast, reliable fashion, with minimal downtime and the option to rollback.  Security Hardening: In a shared Cloud environment customers will be extremely security-aware to protect their data and services. The hardening of the OS installation shall happen with minimum effort to the highest sensible levels.  HA Platform: The OS must support a HighAvailability framework enabling customers to failover/scale services among many VMs. This allows customers to do maintenance on one node while the other one is servicing requests, and respectively keep the services running in case of an outage.  Integration with HardWare and HyperVisor: The Operating System must be aware of the HW and HV it is running on, closely cooperating with both in case of resource reconfiguration, maintenance, migration of VMs.  License costs: You know the drill. The lower the stacking costs the happier your procurement will be.  And again, we got this far. We have some important requirements to fulfill. Allow me to introduce Solaris 11, in case you were not familiar with it yet - and also, allow me to show you how they fulfill these requirements. SOLARIS 11 ZFS Zones Database and Java integration Network Virtualization  DTrace  CryptoFramework  Hardened by default How Oracle Solaris 11 fulfills these requirements is:   Application separation: Solaris comes with the built-in virtualization technology, Solaris Zones. Zones are an overhead-free userspace OS-virtualization, separating application environments with dedicated userspace, dynamic resource management, dedicated users, separate software packages, their own network configuration, while avoiding the virtualization trap: there's no performance loss, for they are userspace containers.With Solaris virtualization you can effectively overcommit the servers' CPU resources, consolidating more than usual, achieving a higher server utilization. The higher the servers are utilized, the most companies can get for their investment in servers.  Safe Updates: Solaris comes with the BootEnvironments feature, that allows cloning the complete, running OS installation to create bootable fallback images before changes. This feature is completely integrated with Solaris 11's package system too - if one updates the operating system with a simple "pkg update" command, the packaging system creates a clone of the OS setup, and upgrades the packages on the clone instance, without changing anything on the running, live instance. The administrator can then decide when to boot and test the new image. (And of course can fall back to the original one, or any other, previously created one.) Bootenvironments make remote Solaris 11 administration extremely safe.  Security Hardening: Solaris 11 is hardened already in the initial installation state. Just a minimum set of packages is installed, negligible services are disabled. Role Based Access Control (RBAC) is extended and encouraged to use, packages are all cryptographically signed, the cryptoframework is ready to encrypt data. Security hardening is essential in a shared cloud environment.  HA Platform: For Solaris there is a complete HighAvailability Framework, Solaris Cluster, that is a multipurpose cluster framework, binding different Solaris installations into a cluster, failing over services from one clusternode to another one in case one Server/VM would fail, ensuring application availability. It comes with several built-in clusteragents for Oracle DB, MySQL, Zones, Java application servers, Webservers, etc. Solaris cluster keeps your services running even if a single server would fail.  License costs: The cost of running Solaris on Oracle Servers has been reduced to zero. All you need is a support contract for the HW to get the latest and greatest updates, fixes. Do you still pay for OS licenses? Why?  SUMMARY Solaris 11 fulfills the most common IaaS platform requirements for operating systems, with default security hardening, application separation with OS virtualization, a multipurpose Clusterframework, zero license costs, safe updates and tight integration with the Oracle servers.  In the previous post we had a look into the requirements for virtualization for IaaS platforms, and how the LDoms fulfill those.In the next post we will look into the requirements for a shared storage of choice for IaaS platforms.  If questions arise, do not hesitate to comment.   -- charlie 

This is Part III of the series "How to build a private IaaS cloud platform?", elaborating about the platform requirements to face when building a multipurpose Infrastructure as a Service platform.In...

How to build a private IaaS cloud platform? - Part II, Virtualisation

This is Part II of the series "How to build a private IaaS cloud platform?", elaborating about the platform requirements to face when building a multipurpose Infrastructure as a Service platform. In this post we are looking into the requirements the chosen virtualization technology has to fulfill. See the Main Summary Post with links to the other sections (requirements analysis for OS, shared storage, virtualization, etc) here.  Part II: Virtualization So, what are the most important requirements a virtualization technology has to fulfill to be chosen as an IaaS cloud building block? Requirements list: Virtualization performance penalty: Virtualisation technologies are often - and in some cases, legitimately - considered  as performance bottlenecks. Emulating complete virtual machines is not easy to implement, mapping to physical devices, scheduling computing requests from virtual CPUs to physical strands, channeling I/O... The goal is of course to keep the performance loss minimal in a virtualized environment so that your IaaS platform can scale well.  Livemigration between hosts: Physics is bound to have failures. Fault tolerance, power management and server maintainability, requires Virtual Machines to be migrateable in a live fashion, that is, without downtime from one physical host to another one. This must happen in a secure and automatable way.  Dynamic reconfiguration: Of course one wants to adjust the resources of running VMs without downtime, like adding/growing virtual disks or network interfaces, assigning more CPUs/threads and/or RAM. This of course must happen in cooperation of the operating system running in the VM, that is, with OS awareness. Software license boundaries: This one isn't actually a technical requirement, but rather an organisational one. Providing VMs running - for example - an Oracle Database to your customers is a great idea - but customers tend to prefer to be charged only for licenses of the CPUs dedicated to their VMs and not for the whole server. Not all virtualisation technologies are approved license boundaries.  Virtualisation costs: Of course the less costs come up for adding a virtualisation technology to your stack, the happier the management level will be.  Alright, we got this far. We have some important requirements to fulfill. Now let me introduce you to Oracle VM for SPARC (former LDoms) in case you were not familiar with them - and also, allow me to show you how they fulfill these requirements. Oracle VM for SPARC How OVM for SPARC fulfills these requirements: Avoiding virtualization performance penalty: OVM for SPARC, aka LDoms are not an emulating type virtualisation, but a partitioning type. The hypervisor actually runs in the firmware of the server. That is, on SPARC servers with the S3 CPU core (T4...) systemadministrators do not emulate servers/VMs but they carve up the physical serverhardware in separate partitions, all with native CPU and Memory dedication, and - depending on the configuration - even with physical I/O. That is, isolated Operating Systems installed in the separate LDoms run natively on the hardware. No virtualization performance loss. Conclusion: With OVM for SPARC system engineers get the complete performance of the serverhardware, which makes it an ideal target for heavily virtualized environments, like IaaS cloud platforms.  Livemigration between physical hosts: Livemigration is of course taken for granted - but what makes this implementation special is rather the security of the migration and the integration of automatisms.  LDoms have Secure Livemigration, because the traffic between the physical nodes is secured via encryption. Not only it is encrypted, but also encrypted with HW support, with the famous HW-cryptoengine of the SPARC S3 CPU cores.  The livemigration feature can be integrated with Solaris Cluster, enabling LDoms to be part of a multipurpose HA-platform. Another option is to utilize EM Ops Center, the datacenter infrastructure management tool originating from Sun Microsystems, that can remotely manage different virtualization technologies, including LDoms with livemigration.  Conclusion: Security, HighAvailability and Managability distinguis LDoms for Infrastructure as a Service platforms.  Dynamic reconfiguration: Systemadministrators can add (and remove) RAM, CPU, network interfaces, disks, PCI slots, cryptounits to running LDoms. The OS, of course must be able to handle inflight HW configuration changes, but Solaris did that since the '90s. Dynamic reconfiguration adds up to availability of running services immensely.  Software license boundaries: LDoms are a license hard-partitioning approved virtualisation technology and can effectively limite the necessary number of licenses to the configured VMs. This enables platform administrators to easily implement license chargeback mechanisms for their customers.  Virtualisation costs: Using LDoms does not need to be licensed. They are simply a feature of the Servers. If your goal is to minimize virtualization costs, here you go. Minimized to zero.  SUMMARY LDoms fulfill the most common IaaS platform requirements for virtualization technologies with native Server performance, secure livemigration, dynamic reconfiguration, approved software license boundaries and zero virtualisation costs.  In the previous post we had a look into the requirements for servers for IaaS platforms, and how the SPARC-T Servers fulfill those.In the next post we will look into the requirements for an Operating System of choice for IaaS platforms.  If questions arise, do not hesitate to comment.   -- charlie 

This is Part II of the series "How to build a private IaaS cloud platform?", elaborating about the platform requirements to face when building a multipurpose Infrastructure as a Service platform.In...

How to build a private IaaS cloud platform? - Part I, Servers

This is Part I of the series "How to build a private IaaS cloud platform?", elaborating about the platform requirements to face when building a multipurpose Infrastructure as a Service platform. In this post we are looking into the requirements the chosen servers have to fulfill. See the Main Summary Post with links to the other sections (requirements analysis for OS, shared storage, virtualization, etc) here.  Part I: CPU and Server What are the requirements a server and its CPU architecture need to fulfill to be chosen as an IaaS cloud building block?  Requirements list: CPU Threading: Your customers are going to want to run a lot of different applications on the cloud platform we're building, all with different preferences for threading capabilities. Some will require strong multithreading and throughput, others will prefer high single-thread performance, and there are those that need both these (e.g. the Oracle Database). That is, the choice of CPU can't simply concentrate on one profile, will have to fulfill both, to be considered a general purpose cloud architecture.  Data Encryption Overhead: Your customers will be concerned by the fact that within your cloud eventually other projects, other customers have their applications running too. The question about data security will rise inevitably: "Is my data secure in your cloud?". Sure it is. As long as it is encrypted, you're on the safe side. I hear you say: "But encryption is so expensive in terms of performance!". Right. Usually it is. That is, another requirement for your platform: Keep the encryption overhead as small as possible. Migration risks: Many of your customers are running a large number of servers, legacy applications on legacy setups on legacy platforms, and are cautious about platform upgrades for risking breaking their applications' functionality. To provide a general purpose consolidation platform, you have to minimize the migration risk. Virtualisation: Obviously we're going to run virtual machines on the platform, and the server/CPU of choice must support this intention. Virtualisation requirements alone need an own discussion and will be discussed in the next post.  Consolidation power: Any server worth its salt to be considered as a consolidation platform must be powerful enough to run a plethora of virtual machines.  Alright, we got this far. We have some important requirements to fulfill. Now let me introduce you to the SPARC T4 Servers in case you were not familiar with them - and also, allow me to show you how they fulfill these requirements. SPARC T4 CPU and Servers: How the T4 fulfills these requirements: CPU Threading: The SPARC T4 CPUs all have 8 cores. Each of these 8 cores run 8 threads paralelly, sharing the resources of that CPU core. That's right, that's 64 threads per core. That is, this is one of the modes it can run in. It also capable to run one thread, dedicating it all the power of the complete CPU core. That is, the T4 CPUs provide both single-thread and throughput performance. The best thing about this? The scheduler switches between the modes on-the-fly, dynamically, that is, the switch happens in operation, transparently. We call this dynamic threading. Conclusion: the T4 CPUs fulfill both singlethread and throughput requirements, making them an excellent general-purpose platform and a great choice for IaaS implementations.  Data Encryption Overhead: The T4 CPUs got hardware-encryption engines within their CPUcores. These are actual encryption- and hashing algorithms cast into silicon. That's right, math in metal(loid). Solaris can use these encryption engines to encrypt data at hardware speed, without having to calculate everything in software, thus sparing CPU cycles. It won't get faster than that. Encrypt your Databases (TDE), filesystems (ZFS), SSL traffic, Java application data (JCE), etc. Data safety without encryption overhead, at your fingertips. Conclusion: The T4 Servers with their zero-overhead encryption are the right choice for building multitenant platforms, like IaaS environments. Migration risk: The CPU in question is a SPARC CPU. The SPARC Instruction Set has been only extended over time, that is, if you had your application running on an earlier SPARC CPU then the operation on the newer generation SPARC CPUs is guaranteed. You can of course recompile it to benefit from the new CPUfeatures, but that isn't a requirement to run the application. Also, for the binary compatibility regarding Solaris, see the Oracle Solaris Guarantee Program.Conclusion: The application migration between legacy SPARC servers to current T4 servers is minimized.  Virtualisation: the SPARC T4 servers carry a unique virtualisation technology in their firmware that allows you to carve up the boxes into several virtual machines called logical domains - I will dedicate these a complete followup post.  Consolidation power: In a 4-Socket T4 Server (the T4-4) there are 4 CPUs, each with 8 CPU cores that can run 8 threads. That is 4x8x8 == 256 threads that can be executed parallelly. How many of your average servers can you consolidate on a box like this, let's say, with 2TB RAM?Conclusion: there is a lot of power under the engine hood of the T4 servers to offer for serverconsolidation projects.  SUMMARY:  The T4 servers fulfill the most common Infrastructure as a Service platform requirements with dynamic threading characteristics to serve all kinds of application performancerequirements, built-in virtualization for multitenancy, HW-encryption for data safety, keeping migration risk at a minimum with SPARC backward compatibility, and a lot of computing power and throughput for platform consolidation.  In the next blogpost we will look into the virtualization technology for the platform, its requirements to fulfill and how it is included - OVM for SPARC, aka LDoms. -- charlie 

This is Part I of the series "How to build a private IaaS cloud platform?", elaborating about the platform requirements to face when building a multipurpose Infrastructure as a Service platform.In...

How to build a private IaaS cloud platform? - Main Post

I had the chance to give talks about evaluating the requirements for, and building private IaaS cloud platforms. Often I was explained the following situation: A corporate manager went to OpenWorld, a local Oracle Day or a Virtualization Day, and has seen everywhere that cloud platforms are the future. As a good businessman, he wants to benefit from the new wave and wants you, the system administrator/datacenter engineer/platform architecht to build an Infrastructure as a Service platform to provide virtual environments for your internal or external customers. What do you do? Allow me to point out some of the requirements and considerations.  The cloud building blocks  Thinking outside of the box and on quite a high-level, you are going to need:  Servers Virtualisation Operating System Shared Storage  A central management tool  All these layers are bound to fulfill some requirements to enable a successful cloud platform. Let's see at each what the functional requirements are to choose the right product for each:  The blogpost has been broken down into a series of 5 smaller sections for readabilitly. This is the main post, linking to all the five sub-posts.   Part I - Servers https://blogs.oracle.com/orasysat/entry/how_to_build_a_private  Part II - Virtualisation https://blogs.oracle.com/orasysat/entry/how_to_build_a_private2  Part III - OS  https://blogs.oracle.com/orasysat/entry/how_to_build_a_private3  Part IV - Shared Storage https://blogs.oracle.com/orasysat/entry/how_to_build_a_private4  Part V - Platform Management  Part VI - Putting it all together: The WhitePaper Update, 03.01.2103: Link to Servers post added Update, 09.01.2103: Link to Virtualization post addedUpdate, 05.02.2103: Link to Operating System post addedUpdate, 27.02.2103: Link to Shared Storage post added

I had the chance to give talks about evaluating the requirements for, and building private IaaS cloud platforms. Often I was explained the following situation: A corporate manager went to OpenWorld, a...

What's new in Solaris 11.1?

Solaris 11.1 is released. This is the first release update since Solaris 11 11/11, the versioning has been changed from MM/YY style to 11.1 highlighting that this is Solaris 11 Update 1.  Solaris 11 itself has been great. What's new in Solaris 11.1? Allow me to pick some new features from the What's New PDF that can be found in the official Oracle Solaris 11.1 Documentation. The updates are very numerous, I really can't include all.  I. New AI Automated Installer RBAC profiles have been introduced to enable delegation of installation tasks. II. The interactive installer now supports installing the OS to iSCSI targets. III. ASR (Auto Service Request) and OCM (Oracle Configuration Manager) have been enabled by default to proactively provide support information and create service requests to speed up support processes. This is optional and can be disabled but helps a lot in supportcases. For further information, see: http://oracle.com/goto/solarisautoreg IV. The new command svcbundle helps you to create SMF manifests without having to struggle with XML editing. (btw, do you know the interactive editprop subcommand in svccfg? The listprop/setprop subcommands are great for scripting and automating, but for an interactive property editing session try, for example, this: svccfg -s svc:/application/pkg/system-repository:default editprop )  V. pfedit: Ever wondered how to delegate editing permissions to certain files? It is well known "sudo /usr/bin/vi /etc/hosts" is not the right way, for sudo elevates the complete vi process to admin levels, and the user can "break" out of the session as root with simply starting a shell from that vi. Now, the new pfedit command provides a solution exactly to this challenge - an auditable, secure, per-user configurable editing possibility. See the pfedit man page for examples.   VI. rsyslog, the popular logging daemon (filters, SSL, formattable output, SQL collect...) has been included in Solaris 11.1 as an alternative to syslog.  VII: Zones: Solaris Zones - as a major Solaris differentiator - got lots of love in terms of new features: ZOSS - Zones on Shared Storage: Placing your zones to shared storage (FC, iSCSI) has never been this easy - via zonecfg.  parallell updates - with S11's bootenvironments updating zones was no problem and meant no downtime anyway, but still, now you can update them parallelly, a way faster update action if you are running a large number of zones. This is like parallell patching in Solaris 10, but with all the IPS/ZFS/S11 goodness.  per-zone fstype statistics: Running zones on a shared filesystems complicate the I/O debugging, since ZFS collects all the random writes and delivers them sequentially to boost performance. Now, over kstat you can find out which zone's I/O has an impact on the other ones, see the examples in the documentation: http://docs.oracle.com/cd/E26502_01/html/E29024/gmheh.html#scrolltoc Zones got RDSv3 protocol support for InfiniBand, and IPoIB support with Crossbow's anet (automatic vnic creation) feature.  NUMA I/O support for Zones: customers can now determine the NUMA I/O topology of the system from within zones.  VIII: Security got a lot of attention too:  Automated security/audit reporting, with builtin reporting templates e.g. for PCI (payment card industry) audits.  PAM is now configureable on a per-user basis instead of system wide, allowing different authentication requirements for different users  SSH in Solaris 11.1 now supports running in FIPS 140-2 mode, that is, in a U.S. government security accredited fashion.  SHA512/224 and SHA512/256 cryptographic hash functions are implemented in a FIPS-compliant way - and on a T4 implemented in silicon! That is, goverment-approved cryptography at HW-speed.  Generally, Solaris is currently under evaluation to be both FIPS and Common Criteria certified.  IX. Networking, as one of the core strengths of Solaris 11, has been extended with:  Data Center Bridging (DCB) - not only setups where network and storage share the same fabric (FCoE, anyone?) can have Quality-of-Service requirements. DCB enables peers to distinguish traffic based on priorities. Your NICs have to support DCB, see the documentation, and additional information on Wikipedia. DataLink MultiPathing, DLMP, enables link aggregation to span across multiple switches, even between those of different vendors. But there are essential differences to the good old bandwidth-aggregating LACP, see the documentation: http://docs.oracle.com/cd/E26502_01/html/E28993/gmdlu.html#scrolltoc VNIC live migration is now supported from one physical NIC to another on-the-fly  X. Data management:  FedFS, (Federated FileSystem) is new, it relies on Solaris 11's NFS referring mechanism to join separate shares of different NFS servers into a single filesystem namespace. The referring system has been there since S11 11/11, in Solaris 11.1 FedFS uses a LDAP - as the one global nameservice to bind them all.  The iSCSI initiator now uses the T4 CPU's HW-implemented CRC32 algorithm - thus improving iSCSI throughput while reducing CPU utilization on a T4 Storage locking improvements are now RAC aware, speeding up throughput with better locking-communication between nodes up to 20%!  XI: Kernel performance optimizations: The new Virtual Memory subsystem ("VM2") scales now to 100+ TB Memory ranges.  The memory predictor monitors large memory page usage, and adjust memory page sizes to applications' needs OSM, the Optimized Shared Memory allows Oracle DBs' SGA to be resized online XII: The Power Aware Dispatcher in now by default enabled, reducing power consumption of idle CPUs. Also, the LDoms' Power Management policies and the poweradm settings in Solaris 11 OS will cooperate. XIII: x86 boot: upgrade to the (Grand Unified Bootloader) GRUB2. Because grub2 differs in the configuration syntactically from grub1, one shall not edit the new grub configuration (grub.cfg) but use the new bootadm features to update it. GRUB2 adds UEFI support and also support for disks over 2TB. XIV: Improved viewing of per-CPU statistics of mpstat. This one might seem of less importance at first, but nowadays having better sorting/filtering possibilities on a periodically updated mpstat output of 256+ vCPUs can be a blessing. XV: Support for Solaris Cluster 4.1: The What's New document doesn't actually mention this one, since OSC 4.1 has not been released at the time 11.1 was. But since then it is available, and it requires Solaris 11.1. And it's only a "pkg update" away. ...aand I seriously need to stop here. There's a lot I missed, Edge Virtual Bridging, lofi tuning, ZFS sharing and crypto enhancements, USB3.0, pulseaudio, trusted extensions updates, etc - but if I mention all those then I effectively copy the What's New document. Which I recommend reading now anyway, it is a great extract of the 300+ new projects and RFE-followups in S11.1. And this blogpost is a summary of that extract.  For closing words, allow me to come back to Request For Enhancements, RFEs. Any customer can request features. Open up a Support Request, explain that this is an RFE, describe the feature you/your company desires to have in S11 implemented. The more SRs are collected for an RFE, the more chance it's got to get implemented. Feel free to provide feedback about the product, as well as about the Solaris 11.1 Documentation using the "Feedback" button there. Both the Solaris engineers and the documentation writers are eager to hear your input.Feel free to comment about this post too. Except that it's too long ;)  wbr,charlie

Solaris 11.1 is released. This is the first release update since Solaris 11 11/11, the versioning has been changed from MM/YY style to 11.1 highlighting that this is Solaris 11 Update 1.  Solaris 11...

Clustering Basics and Challenges

For upcoming posts it seemed to be a good idea to dedicate some time for cluster basic concepts and theory. This post misses a lot of details that would explode the articlesize, should you have questions, do not hesitate to ask them in the comments.  The goal here is to get some concepts straight. I can't promise to give you an overall complete definitions of cluster, cluster agent, quorum, voting, fencing, split brain condition, so the following is more of an explanation. Here we go. -------- Cluster, HA, failover, switchover, scalability -------- An attempted definition of a Cluster: A cluster is a set (2+) server nodes dedicated to keep application services alive, communicating through the cluster software/framework with eachother, test and probe health status of servernodes/services and with quorum based decisions and with switchover/failover techniques keep the application services running on them available. That is, should a node that runs a service unexpectedly lose functionality/connection, the other ones would take over the and run the services, so that availability is guaranteed. To provide availability while strictly sticking to a consistent clusterconfiguration is the main goal of a cluster.  At this point we have to add that this defines a HA-cluster, a High-Availability cluster, where the clusternodes are planned to run the services in an active-standby, or failover fashion. An example could be a single instance database. Some applications can be run in a distributed or scalable fashion. In the latter case instances of the application run actively on separate clusternodes serving servicerequests simultaneously. An example for this version could be a webserver that forwards connection requests to many backend servers in a round-robin way. Or a database running in active-active RAC setup.  -------- Cluster arhitecture, interconnect, topologies -------- Now, what is a cluster made of? Servers, right. These servers (the clusternodes) need to communicate. This of course happens over the network, usually over dedicated network interfaces interconnecting all the clusternodes. These connection are called interconnects.How many clusternodes are in a cluster? There are different cluster topologies. The most simple one is a clustered pair topology, involving only two clusternodes:  There are several more topologies, clicking the image above will take you to the relevant documentation. Also, to answer the question Solaris Cluster allows you to run up to 16 servers in a cluster. Where shall these clusternodes be placed? A very important question. The right answer is: It depends on what you plan to achieve with the cluster. Do you plan to avoid only a server outage? Then you can place them right next to eachother in the datacenter. Do you need to avoid DataCenter outage? In that case of course you should place them at least in different fire zones. Or in two geographically distant DataCenters to avoid disasters like floods, large-scale fires or power outages. We call this a stretched- or campus cluster, the clusternodes being several kilometers away from eachother. To cover really large distances, you probably need to move to a GeoCluster, which is a different kind of animal.  What is a geocluster? A Geographic Cluster in Solaris Cluster terms is actually a metacluster between two, separate (locally-HA) clusters.  -------- Cluster resource types, agents, resources, resource groups -------- So how does the cluster manage my applications? The cluster needs to start, stop and probe your applications. If you application runs, the cluster needs to check regularly if the application state is healthy, does it respond over the network, does it have all the processes running, etc. This is called probing. If the cluster deems the application is in a faulty state, then it can try to restart it locally or decide to switch (stop on node A, start on node B) the service. Starting, stopping and probing are the three actions that a cluster agent does. There are many different kinds of agents included in Solaris Cluster, but you can build your own too. Examples are an agent that manages (mounts, moves) ZFS filesystems, or the Oracle DB HA agent that cares about the database, or an agent that moves a floating IP address between nodes. There are lots of other agents included for Apache, Tomcat, MySQL, Oracle DB, Oracle Weblogic, Zones, LDoms, NFS, DNS, etc.We also need to clarify the difference between a cluster resource and the cluster resource group.A cluster resource is something that is managed by a cluster agent. Cluster resource types are included in Solaris cluster (see above, e.g. HAStoragePlus, HA-Oracle, LogicalHost). You can group cluster resources into cluster resourcegroups, and switch these groups together from one node to another. To stick to the example above, to move an Oracle DB service from one node to another, you have to switch the group between nodes, and the agents of the cluster resources in the group will do the following:  On node A Shut down the DB Unconfigure the LogicalHost IP the DB Listener listens on unmount the filesystem   Then, on node B: mount the FS configure the IP  startup the DB -------- Voting, Quorum, Split Brain Condition, Fencing, Amnesia -------- How do the clusternodes agree upon their action? How do they decide which node runs what services? Another important question. Running a cluster is a strictly democratic thing.Every node has votes, and you need the majority of votes to have the deciding power. Now, this is usually no problem, clusternodes think very much all alike. Still, every action needs to be governed upon in a productive system, and has to be agreed upon. Agreeing is easy as long as the clusternodes all behave and talk to eachother over the interconnect. But if the interconnect is gone/down, this all gets tricky and confusing. Clusternodes think like this: "My job is to run these services. The other node does not answer my interconnect communication, it must be down. I'd better take control and run the services!". The problem is, as I have already mentioned, clusternodes very much think alike. If the interconnect is gone, they all assume the other node is down, and they all want to mount the data backend, enable the IP and run the database. Double IPs, double mounts, double DB instances - now that is trouble. Also, in a 2-node cluster they both have only 50% of the votes, that is, they themselves alone are not allowed to run a cluster.  This is where you need a quorum device. According to Wikipedia, the "requirement for a quorum is protection against totally unrepresentative action in the name of the body by an unduly small number of persons.". They need additional votes to run the cluster. For this requirement a 2-node cluster needs a quorum device or a quorum server. If the interconnect is gone, (this is what we call a split brain condition) both nodes start to race and try to reserve the quorum device to themselves. They do this, because the quorum device bears an additional vote, that could ensure majority (50% +1). The one that manages to lock the quorum device (e.g. if it's an FC LUN, it SCSI reserves it) wins the right to build/run a cluster, the other one - realizing he was late - panics/reboots to ensure the cluster config stays consistent.  Losing the interconnect isn't only endangering the availability of services, but it also endangers the cluster configuration consistence. Just imagine node A being down and during that the cluster configuration changes. Now node B goes down, and node A comes up. It isn't uptodate about the cluster configuration's changes so it will refuse to start a cluster, since that would lead to cluster amnesia, that is the cluster had some changes, but now runs with an older cluster configuration repository state, that is it's like it forgot about the changes.  Also, to ensure application data consistence, the clusternode that wins the race makes sure that a server that isn't part of or can't currently join the cluster can access the devices. This procedure is called fencing. This usually happens to storage LUNs via SCSI reservation.  Now, another important question: Where do I place the quorum disk?  Imagine having two sites, two separate datacenters, one in the north of the city and the other one in the south part of it. You run a stretched cluster in the clustered pair topology. Where do you place the quorum disk/server? If you put it into the north DC, and that gets hit by a meteor, you lose one clusternode, which isn't a problem, but you also lose your quorum, and the south clusternode can't keep the cluster running lacking the votes. This problem can't be solved with two sites and a campus cluster. You will need a third site to either place the quorum server to, or a third clusternode. Otherwise, lacking majority, if you lose the site that had your quorum, you lose the cluster. Okay, we covered the very basics. We haven't talked about virtualization support, CCR, ClusterFilesystems, DID devices, affinities, storage-replication, management tools, upgrade procedures - should those be interesting for you, let me know in the comments, along with any other questions. Given enough demand I'd be glad to write a followup post too. Now I really want to move on to the second part in the series: ClusterInstallation.  Oh, as for additional source of information, I recommend the documentation: http://docs.oracle.com/cd/E23623_01/index.html, and the OTN Oracle Solaris Cluster site: http://www.oracle.com/technetwork/server-storage/solaris-cluster/index.html

For upcoming posts it seemed to be a good idea to dedicate some time for cluster basic concepts and theory. This post misses a lot of details that would explode the articlesize, should you...

10 tape technology features that make you go hmm.

A week ago an Oracle/StorageTek Tape Specialist, Christian Vanden Balck, visited Vienna, and agreed to visit customers to do techtalks and update them about the technology boom going around tape. I had the privilege to attend some of his sessions and noted the information and features that took the customers by surprise and made them think. Allow me to share the top 10: I. StorageTek as a brand: StorageTek is one of he strongest names in the Tape field. The brand itself was valued so much by customers that even after Sun Microsystems acquiring StorageTek and the Oracle acquiring Sun the brand lives on with all the Oracle tapelibraries are officially branded StorageTek.See http://www.oracle.com/us/products/servers-storage/storage/tape-storage/overview/index.html II. Disk information density limitations: Disk technology struggles with information density. You haven't seen the disk sizes exploding lately, have you? That's partly because there are physical limits on a disk platter. The size is given, the number of platters is limited, they just can't grow, and are running out of physical area to write to. Now, in a T10000C tape cartridge we have over 1000m long tape. There you go, you have got your physical space and don't need to stuff all that data crammed together. You can write in a reliable pattern, and have space to grow too. III. Oracle has a market share of 62% worldwide in recording head manufacturing. That's right. If you are running LTO drives, with a good chance you rely on StorageTek production. That's two out of three LTO recording heads produced worldwide.  IV. You can store 1 Exabyte data in a single tape library. Yes, an Exabyte. That is 1000 Petabytes. Or, a million Terabytes. A thousand million GigaBytes. You can store that in a stacked StorageTek SL8500 tapelibrary. In one SL8500 you can put 10.000 T10000C cartridges, that store 10TB data (compressed). You can stack 10 of these SL8500s together. Boom. 1000.000 TB.(n.b.: stacking means interconnecting the libraries. Yes, cartridges are moved between the stacked libraries automatically.)  V. EMC: 'Tape doesn't suck after all. We moved on.': Do you remember the infamous 'Tape sucks, move on' Datadomain slogan? Of course they had to put it that way, having only had disk products. But here's a fun fact: on the EMCWorld 2012 there was a major presence of a Tape-tech company - EMC, in a sudden burst of sanity is embracing tape again. VI. The miraculous T10000C: Oracle StorageTek has developed an enterprise-grade tapedrive and cartridge, the T10000C. With awesome numbers: The Cartridge: Native 5TB capacity, 10TB with compression Over a kilometer long tape within the cartridge. And it's locked when unmounted, no rattling of your data.  Replaced the metalparticles datalayer with BaFe (bariumferrite) - metalparticles lose around 7% of magnetism within 30 days. BaFe does not. Yes we employ solid-state physicists doing R&D on demagnetisation in our labs. Can be partitioned, storage tiering within the cartridge!  The Drive: 2GB Cache Encryption implemented in HW - no performance hit 252 MB/s native sustained data rate, beats disk technology by far. Not to mention peak throughput.  Leading the tape while never touching the data side of it, protecting your data physically too Data integritiy checking (CRC recalculation) on tape within the drive without having to read it back to the server reordering data from tape-order, delivering it back in application-order  writing 32 tracks at once, reading them back for CRC check at once VII. You only use 20% of your data on a regular basis.The rest 80% is just lying around for years. On continuously spinning disks. Doubly consuming energy (power+cooling), blocking diskstorage capacity. There is a solution called SAM (Storage Archive Manager) that provides you a filesystem unifying disk and tape, moving data on-demand and for clients transparently between the different storage tiers. You can share these filesystems with NFS or CIFS for clients, and enjoy the low TCO of tape. Tapes don't spin. They sit quietly in their slots, storing 10TB data, using no energy, producing no heat, automounted when a client accesses their data.See: http://www.oracle.com/us/products/servers-storage/storage/storage-software/storage-archive-manager/overview/index.html VIII. HW supported for up to two decades: Did you know that the original PowderHorn library was released in '93 and has been only discontinued in 2010? That is nearly two decades of supported operation. Tape libraries are - just like the data carrying on tapecartridges - built for longevity. Oh, and the T10000C cartridge has 30-year archival life for long-term retention.  IX. Tape is easy to manage: Have you heard of Tape Storage Analytics? It is a central graphical tool to summarize, monitor, analyze dataflow, health and performance of drives and libraries, see: http://www.oracle.com/us/products/servers-storage/storage/tape-storage/tape-analytics/overview/index.html X. The next generation: The T10000B drives were able to reuse the T10000A cartridges and write on them even more data. On the same cartridges. We call this investment protection, and this is very important for Oracle for the future too. We usually support two generations of cartridges together. The current drive is a T10000C. (...I know I promised to enlist 10, but I got still two more I really want to mention. Allow me to work around the problem: ) X++. The TallBots, the robots moving around the cartridges in the StorageTek library from tapeslots to the drives are cableless. Cables, belts, chains running to moving parts in a library cause maintenance downtimes. So StorageTek eliminated them. The TallBots get power, commands, even firmwareupgrades through the rails they are running on. Also, the TallBots don't just hook'n'pull the tapes out of their slots, they actually grip'n'lift them out. No friction, no scratches, no zillion little plastic particles floating around in the library, in the drives, on your data. (X++)++: Tape beats SSDs and Disks. In terms of throughput (252 MB/s), in terms of TCO: disks cause around 290x more power and cooling, in terms of capacity: 10TB on a single media and soon more.  So... do you need to store large amounts of data? Are you legally bound to archive it for dozens of years? Would you benefit from automatic storage tiering? Have you got large mediachunks to be streamed at times? Have you got power and cooling issues in the growing datacenters? Do you find EMC's 180° turn of tape attitude interesting, but appreciate it at the same time? With all that, you aren't alone. The most data on this planet is stored on tape. Tape is coming. Big time.

A week ago an Oracle/StorageTek Tape Specialist, Christian Vanden Balck, visited Vienna, and agreed to visit customers to do techtalks and update them about the technology boom going around tape.I had...

Summary of the Solaris 11 webcast's livechat QnA session

This is a followup post to the previous summary on the "What's new with Solaris 11 since the launch" webcast. That webcast has had a chatroom for a live Questions and Answers session running. I went through the archive of those and compiled a list of some of the (IMHO) most relevant and most frequently asked questions, I'd like to share. This is the first part, covering the QnA of Session I and II of the webcast, in a followup post we can have a look of the rest of the sessions if required - let me know in the comments. Also, should you have questions, as usual, feel free to ask those there, too.  ...and here come the answered questions:  When will Exadata be based on Solaris in place of Oracle Enterprise Linux?Exadata offers both Solaris 11 or Oracle Enterprise Linux.  The choice can be made at deployment time based on your OS needs.What are all other benefits and futures avilable in solaris 11 (cloud O.S.) compared to cloud based Red Hat Linux and Windows?suggest you check out our cloud white paper for a view of this. Also the OTN Solaris 11 page has some good articles. Here are the links:  http://www.oracle.com/technetwork/server-storage/solaris11/documentation/o11-106-sol11-cloud-501066.pdf http://www.oracle.com/technetwork/server-storage/solaris11/overview/index.htmlWill 11.1 have a more complete IPS respository for Oracle and FOSS software?Yes, we are adding additional packages to the various package repositories. Since Solaris 11 was launched, both the Oracle Solaris Studio tools as well as Oracle Solaris Cluster have been made available along with numerous new FOSS packages. We will continue to be adding additional Oracle products and open source packages in the future. Will Exadata be based on Sparc in place of intel-amd x86 in next future ?We can't publically discuss futures, but we actually have a SPARC version of Exadata today, it's called SuperCluster, this is such a powerfull multipurpose system that it actually have multiple personalities built into one system: Exadata, Exalogic, and it can be a general purpose platform if you want. Have I understood this right? Livepatching KSplice-style is coming to Solaris 11 too?We're looking at that for certain types of Solaris patches in the future.Will there be a security framework like SST/JASS for Solaris 11?We can't talk about the future projects on a public forum, but we recognize the need for SST/JASS and want to address this as soon as possible. On the other side there are a whole bunch of "best practices" that are now embedded into Solaris 11 by default, so out of the box Solaris 11 should already address part of what SST/JASS gave you. (For example we did a lot of work on improving the auditing performance so that we can now have it turned on by default). On x86 can install VirtualBox in a Zone and use that to host other OSes.Yes, this was one of the first things we made sure would work when we acquired VirtualBox when we were still Sun Microsystems. If I have a Solaris 11 Control Domain on a T-series, can I run a Solaris 10 Ldom with Solaris 8 branded containers?Yes, you can.Is Oracle Solaris free or do we need to purchase?Solaris is free, the entitlement to run it comes either with a Sun system (new or historical) or for 3rd party systems the entitlement comes with a support contract. Note that for production use you will be expected to get a support contract. If you don't want to use the Solaris system (Sun or 3rd party) for production use (i.e. development) you can get an OTN license on the Oracle Technical Network website. Will encryption and deduplication both work on a share?This should work at the same time. What approaches does Solaris use to monitor usage?There are many different tools in Solaris to monitor usage. The main ones are the "stats" (vmstat, mpstat, prstat, ...), the kstat interface, and DTrace (to get details you couldn't see before). And then there are layered tools that can interface with these tools (Ops Center, BMC, CA, Tivoli, ...) Apart little-endian, big-endian how is it easy to port Solaris applications on Sparc to x86 and vice-versa ?Very easy. Except for certain hardware specific applications (those that utilize hardware specific drivers), all of the same Oracle Solaris APIs exist for all architectures. Is IPS based patching aware of the fact that zones can reside on ZFS and move from one physical server to another ?IPS is definitely aware of zones and uses ZFS to support boot environments for non-global zones in the same way that's used for the global zone. With respect to moving a zone from one physical server to another, Solaris 11 supports to the same zone attach/deattach method that was introduced in Solaris 10. Is vnic support in Ldoms planned?This is currently being investigated for a future LDOM release. Is it possible with the new patching system to build a system later with the same patch level as a system built a few months earlier?Yes, you can choose/define exactly which version should go to the system and it will always put the same bits in place. The technical answer is that you choose the version of the "entire" package you want on the system and the rest flows from there. Is it in the plans to allow zones to add/remove zpools to running zones dynamically in future updates?Work in this area is currently under investigation. Any plans to realese Solaris 11 source code? i.e. opensolaris?We currently can't comment on publicly releasing the source code. If you need/want this access please let your Oracle account team know. What about VirtualBox and Solaris11 for virtualization?Solaris 11 works great with VirtualBox, as both a client and a host system. Will Oracle DB software eventually be supplied as IPS packages? When?We don't have a date yet but this is actively being worked on. What are the new artifacts in Oracle Solaris 11 than the previous versions?There are quite a few actually. The best start is to look at our "Evaluate Solaris 11" page, and there you also can find a Transition Guide. http://www.oracle.com/technetwork/server-storage/solaris11/overview/evaluate-1530234.html So, this seems just like RedHat's YUM environment?IPS offers certain features beyond those in YUM or other packaging systems. For example, IPS works with ZFS and Solaris Boot Environments to provide a safe environment for software lifecycle management so that changes can be reverted by switching to an older boot environment. With Zones on solaris 11, can I do paravirtualitation?The great thing about zones is you don't *need* paravirtualization. You're making the same direct kernel calls that you would outside of a zone.  It's an incredibly significant performance win over hypervisor-based virtualization. Are zones/containers officially supported to run Oracle Databases?  EBIZ?Hi Calvin, the answer is yes, here is the support matrix for DB:  http://www.oracle.com/technetwork/database/virtualizationmatrix-172995.html I've found some nasty bugs in Solaris 11 (one of which today) that have been fixed in community forks (i.e., Illumos). Will Oracle ever restart collaboration with the community?We continue to work with the community, just not as open on all projects as we did before (For example IPS is an open project) and the source of more than half of the Solaris packages is posted on our opensource websites. I can't comment on what we will do in the future. And with regards to bugs please file them through the support organization and we will get them resolved. Is zpool vdev removal on-the fly now possible ?This issue is actively being investigated although we don't have a date for when this feature will be available. Is pgstat now the official replacement for corestat ?It's intended to provide similar functionality Where are the opensource website?For Oracle Solaris, visit http://www.oracle.com/technetwork/opensource/systems-solaris-1562786.html As a cloud-scale virtualization, is it going to be easier to move zones between machines? maybe even automatic in case of a hardware failure?Hi Gashaw, we already have customers that have implemented what they refer to as "flying zones" that they can move around very easily. They use Solaris Cluster to do this. What about VMware vMotion like feature?We have secure live migration with both Logical Domains on SPARC T series systems, and with Oracle VM on x86 systems. When running Solaris 10/11 on an enterprise server with a lot of zones, what are best practises commands to show the system is running fine? (has enough hardware resources). For example CPU / Memory / I/O / system load. What are the recommended values?For Solaris 11, look into the new zonestat(1M) command that provides a great deal of information about zone utilization. In addition, there is new work underway in providing additional observability in areas such as per-zone file system I/O. Java optimizations done with Solaris 11? For X86 platforms too? Where can I find more detail about this?There is lots of work that go into optimizing Java for Oracle Solaris 10 & 11 on both SPARC and x86. See http://www.oracle.com/technetwork/articles/servers-storage-dev/solarisforjavadevelop-168642.pdf What is meant by "ZFS Shadow Migration"?It's a way to migrate data from another file system to ZFS: http://docs.oracle.com/cd/E23824_01/html/E24456/filesystem-3.html Is flash archive available with S11?Flash archive is not.  There is a procedure for disaster recovery, and we're working on a modern archive-based deployment tool for a future update.  The disaster recovery tool is here: http://www.oracle.com/technetwork/articles/servers-storage-admin/o11-091-sol-dis-recovery-489183.html You can also use Distribution Constructor to build common golden images. Will solaris 11 be available on the ODA soon?The idea's under evaluation -- we'll share your interest with the team. What steps can be taken to ensure that breaches of security are identified quickly?There are a number of tools, including the "bart" tool and "pkg verify" to ensure that software has not been compromised.  Solaris Audit can also be used to detect unauthorized access.  You can also use Immutable Zones to protect against compromise.  There are a wide variety of security tools, and I've covered only a few. What is the relation from solaris to java 7 speed optimization?There is constant work done between the Oracle Solaris and Java teams on performance optimizations. See http://docs.oracle.com/javase/7/docs/technotes/guides/vm/performance-enhancements-7.html for examples. What is the difference in the Solaris 11 installation compared to solaris 10 ? where i can find the document describing basic repository concepts ?The best place to start is: http://www.oracle.com/technetwork/server-storage/solaris11/index.html Hope you found the post useful. For questions, input, requests for the second half of the QnA, please find the comment section below.  -- charlie  

This is a followup post to the previous summary on the "What's new with Solaris 11 since the launch" webcast. That webcast has had a chatroom for a live Questions and Answers session running. I...

Summary: What's new with Solaris 11 since the launch?

There was a great online forum titled: Solaris 11: What's new since the launch? last week, and it has brought quite an amount of update and information about what indeed is going on on the Solaris 11 roadmap and how Solaris interacts and extends other products. I recommend you to watch it (just register, or if you have registered, after providing the registered mailaddress you can re-watch the recorded session.). For the ones lacking the time to watch the videos, allow me to attempt a summary.  Please keep in mind, that this is an extract from the presentation that very much includes the safe harbour statement that says that this presentation is solely intended to outline general product direction, and it represents no commitment whatsoever. That slide is very much is in effect in this blogpost too.  Markus Flierl, VP, Software Development took the first slot: He explained that Solaris 11 has been released early November last year. We are running it already longer in production as an essential part of some of our core products like Exadata, SPARC Supercluster, or the ZFS StorageAppliance.  Parallel to that the SPARC-T4 based systems have pleasantly surprised customers with 5-7x improvement over the expected single thread performance. He also has mentioned that the T5 based systems are already running tests in our labs. He went on talking about the Solaris Cluster 4.0 release, Ops Center 12c the latest version of Sun's integrated datacenter management tool, and Solaris Studio 12.3 the development platform for Solaris have been all updated and shipped supporting Solaris 11. That is, all in all, Oracle is investing and pushing the complete Solaris-related portfolio of former Sun products.  Markus then explained that the Solaris 11 had two different goals:  I. Maintaining the traditional Solaris values/properties: availability, scalability, security and performance II. Introducing the long-avaited innovative technologies like the network virtualization layer, the new virtual memory subsystem, the improved zones functionality, the cryptoframework and ZFS updates, bootenvironments, and so on. As Markus explained we expect Solaris 11 Update 1 to ship around the end of this year, featuring:  A rearchitecture of the virtual memory subsystem is in place that will be able to handle 64-socket servers with up to 64TB of main memory that are going to be shipped in the future. To be able to maintain these huge resources standard VM operations like pagetracking and -placement have been streamlined.  The possibility to do link aggregation across different switches.  The "moving zones" concept will include moving VNIC configurations to the target host too, not only the zonecfg.  Upgrade of installed OS instances takes now even less time, for the update procedure has been tuned further (improved python performance).  A roughly annual update cycle. The Solaris development team works tightly together with the SPARC engineers and the Oracle application developers, essentially making Solaris to the secret sauce that really glues applications and hardware together, utilizing the HW's features to provide services to the applications running on top of Solaris like:  in the T4 CPUs the cryptoengine has been significantly improved, so that the Solaris Cryptoframework can offload encryption functionalities at an ever higher speed to (for example Java) applications (using the Java Crypto Extensions), which in turn do not have to implement these functions and burn CPU cycles in software, but run at HW-encryption speed.  Upcoming CPUs will sport similar possibilities, offloading frequently executed application load into HW, like the Oracle numbers arithmetic acceleration, or data decompression.  The scheduler is being adapted to handle JVM or Database specific workloads targeted to the needs of those. A hotpatching technology similar to KSplice providing runtime upgrades of some running OS elements without downtime is presumably coming to a later Solaris 11 update Java7 Update4 is featuring the Java Mission Control tech, that traces application activities in the JVM, but also ties into Solaris's DTrace framework, essentially providing an end-to-en view on the application performance on the server.  Of course the M4 servers that are already running in our labs are being tested and have been developed with Solaris 11 too. We're trying to align these hardware releases with the software releases.  Technology that is being looked into for upcoming Solaris 11 updates and that is coming up on different products : The ZFS StorageAppliance is about to get an encryption option too  The Exadata Database machine will be enabled to use Solaris's Zones to further improve existing multitenancy capabilities of the box (and Solaris's NUMA I/O tuning has already been the key to achieve 1 Million IOPS on an Exadata box!)  Markus also pointed out that although the Solaris 11 is very much well defined in details, large plans for Solaris 12 are being discussed already.   To wrap it up: one of the interviewed customer leads has defined Solaris customers as being historically very aggressive users of technology. With Solaris 11 many requested features have been implemented for them, while still keeping the well established values of scalability, security, performance that is highly valued by the other half of our customers.  A handy URL to access the recording of the sessions, including the Q&A: http://bit.ly/os11forum2012The public SPARC/Solaris large-scale roadmap: http://bit.ly/sparc_roadmap The slides for Markus's session: http://bit.ly/os11forum2012_keynoteslides I haven't yet summarized the other three sessions, should you like to have that, drop me a comment on this post.  -- charlie  

There was a great online forum titled: Solaris 11: What's new since the launch? last week, and it has brought quite an amount of update and information about what indeed is going on on the Solaris 11...

So, what makes ZFS so cool? (Part I: high level overview)

I have the privilege to do Solaris 11 tech-updates/demos at customers. It always amazes me how much they are amazed by ZFS. Don't get me wrong, ZFS is really cool. But it isn't exactly new technology, it's been around for a while now, the first implementations in 2003, included in Solaris 10 since S10 Update 2 in 2006. Everyone has heard about it being awesome, but every now and then I get the question for details: So tell me, what really makes ZFS so cool?  Let me tell you about it. First and foremost: What were Sun's motivations to go and implement a new data management technology? Let's see.  They had enough of storage capacity limitations of existing filesystems  They were fed up with the complexity and the static nature of the managing data systems  They considered data loss due to silent datacorruption inacceptable  They could not bear the thought of partially written IOs endangering data consistence They wanted consistent rollback functionality to any other previous states They believed tools external to the filesystem were unreliable and an unintegrated way to provide data services They wanted performance through using hybrid storage elements and transparent caching/fetching within the pool.  So, how did they address these issues?  To remove capacity limitations, they made ZFS a 128-bit filesystem. This makes ZFS capable to address 256 quadrillion zettabytes. With this addressspace you could practically store all the digital data ever generated on Earth. That is, you probably never will meet the problem having to create another zpool because the existing one isn't capable to address more storage. [1]  To reduce administration complexity, they did the following: They have moved RAID functionality from external (SVM, VxVM, HW raid controllers) into the zpools They eliminated semi-static logical volume management completely, and defined filesystems with no static definition of size, but as simple, hierarchical management points in the pool.  That is, no need to grow/shrink volumes, for they do not exist, and no need to grow/shrink filesystems, for they are dynamically growing and shrinking in the zpool with the amount of data changing within them. Also, all you need to use is only two commands, zpool and zfs with their intuitively usable subcommands (create/list/destroy/get/...) to manage your data structures.  To avoid data corruption, for they understood that you can't avoid physical bitrot, that is silent datacorruption on the disks, so they have decided to checksum every single block written into the pool. These checksums are controlled at read-time and self-healed from the redundant blocks (mirror, raidz).  To forego partially successful writes ruining data consistence they have implemented ZFS as a transactional filesystem. That is, either a write is completely done, or not at all. Also, changing data happens on a Copy on Write way, that is, reading the relevant blocks, and not modifying the ones being changed, but writing the blocks with changed content to an unused area, leaving the original ones untouched. Both original and changed states are exist at the end of the modification write, then original blocks are marked as freespace (metadata released). To be able to rollback to previous states they snapshot the metadata, and after CoW modification simply not throwing that away. See? Doing snapshots, by not releasing metadata and blocks. It is sometimes easier to snapshot than not to :)  They implemented zfs-internal data services like encryption, deduplication, compression, snapshots, cloning...  To achieve both read- and write performance, they have implemented a hierarchical and configurable cache mechanism, using main memory, L2 cache (even with SSDs) and disks, and autotiering between them.  So if next time anyone asks you about why ZFS is cool, tell them:  amazing storage addressing capability builtin RAID, no need for LVM, dynamic hierarchical filesystems  no datacorruption due to everything is checksummed and there is self-healing transactional and copy-on-write features  snapshots as a natural capability builtin dataservices  hybrid storage pool performance   ...and then of course we didn't yet talk about replication, shadow migration, shares, hierarchical filesystems, delegation, cache policies, dynamical property settings, dynamic striping and autoexpand, online versionupgrades, etc. Should you want me to write about those in a "Part II: a deeper dive" post - let me know in the comments.  wbr,  charlie [1] Although, who knows. Remember working with 180KB "large" floppies? Now my phone has a storage capacity 200.000 times larger than that. My current  phone could replace the data storage needs of a smaller country back in the '80s. According to my quick exponential estimation, if data grows at this speed, then in around a 100 years we will  need something beyond 128 bit filesystems :) 

I have the privilege to do Solaris 11 tech-updates/demos at customers. It always amazes me how much they are amazed by ZFS. Don't get me wrong, ZFS is really cool. But it isn't exactly new technology,...

Great Solaris 10 features paving the way to Solaris 11

What do customers need the most for their UNIX platform's operation? Time. And headcount, obviously. In both of these fields there is usually room for improvement in corporations these times. That is, even though Solaris 11 is out, and it is the greatest Solaris release ever, some customers will probably need time to upgrade their platforms from Solaris 10 - we understand that. We strongly recommend moving to Solaris 11 as soon as you can, but we also know that business goes  first, with service availability hand-in-hand with operations. In this post I list up some of the technologies that are available already in S10, and will ease the move to S11.  First a short disclaimer: Not long ago I was still a system engineer at a customer, running a lot of S10 installations. We have used most of the features mentioned here. I personally have learned a lot from Jörg Möllenkamp's Less Known Solaris Features book (available in PDF on his blog here) and am thankful for him taking the time and effort for documenting and releasing it, it helped a lot. Most of the mentioned features here are in detail discussed there.  All these features are of course part of Solaris, no extra costs, they are already there.  Many customers run Solaris 10 the traditional way, just like we were used to run Solaris for 20 years now, with initscripts, patchsessions in singleuser mode, applications in the global zone, with SVM, UFS, sharing the rootuser - the listed features should improve the administration experience.  Let's get on with the list.   I. LiveUpgrade:Liveupgrade (LU), allows you to create clones of your OS instance  within the same box, while it is up and running. These clones, also called Alternate BootEnvironments can be patched or upgraded (actually changed in any way, like application installation), without a change or downtime of the running system. If you are done with the changes, you can boot your new bootenvironment, and run the patched/changed OS instance there. If the patch/upgrade hasn't worked out, you can reboot back to the original (primary) bootenvironment any time to the unchanged OS instance. This way you reduce downtime, and reduce risk of the updates. It's like a safety net. Back in my days we also used it sometime just to regularly create a complete fallback instance of the OS installation, in case a junior admin removes /usr or something.  Where to read about it? In the above mentioned LKSF book, Section 3, or in the official Solaris 10 documentation in the Live Upgrade section: http://docs.oracle.com/cd/E23823_01/html/E23801/index.html In Solaris 11 LU has been enhanced and is now called bootenvironments. II. Zones This virtualisation method is the most lightweight kind - zones are about userspace separation of your applications, an OS virtualisation, with their separate software packages, daemons, users, IP addresses. They are also well integrated with resource management, that is you can separate now different environments having them in different zones, and you can also guarantee them a minimum amount of CPU power. I have written a short post comparing zones to LDOMs. The conclusion there was: even if it is the only application that will run on that server, put it in a zone. Zones integrate with Solaris Cluster too. Back in my systemadministrator days we have defined them as standard technology, and every new application went into zones.  See the Zones section in the Solaris 10 documentation: http://docs.oracle.com/cd/E18752_01/html/817-1592/index.htmlZones have been improved in Solaris 11, and you can run Solaris10-branded-zones on top of Solaris 11.  III. Resource Management RM  is about ensuring that the resource consumption of a workload (presumably in a zone) does not hinder or interfere with the resource consumption of another one (presumably in another zone or project). The most obvious example is running your application environments in zones, running the zones themselves in a cpupool and the global zone in a different one. In the resourcepool of the zones you can then enable the FairShareScheduler practically guaranteeing to each zone a minimum set of cpushares while allowing them to consume more as long as the cpupool isn't 100% utilized. With this straightforward configuration you reserved some power in the global zone, and practically overcomitted your resources of the cpupool running the zones.  For examples, see the documentation about Resource Pools: http://docs.oracle.com/cd/E18752_01/html/817-1592/rmpool-1.html Using RM is a good way to prepare for consolidation projects.  IV. ZFS ZFS is the last word in filesystems. And volume management. And RAID. These three all in one. No more hardpartitioning your disks for SVM, no more static sized UFS filesystems. ZFS administration is very simple (the two main commands being zfs and zpool - see their intuitive subcommands) and very dynamic. ZFS is supported on Solaris 10 as root filesystem - if you convert your system to ZFS, you can have several LiveUpgrade (see above) Alternate BooteEnvironments within the root-zpool (usually named rpool) based on ZFS snapshots/clones, which are differential, that is only the differences between two environments are written to disk. You can move from UFS to ZFS with LiveUpgrade too. ZFS also provides lots of features, dynamically changeable recordsizes, quota and reservation, snapshots, dumps (zfs send/receive), tuneable I/O performance, and a lot more.  ZFS is the only supported root-filesystem on Solaris 11, moving to this technology on Solaris 10 already will ease your way to S11. For documentation, see the ZFS Administration Guide: http://docs.oracle.com/cd/E23823_01/html/819-5461/index.html V. SMF The Service Management Facility allows you to define services running your OS and running on your OS. These services implement dependency between serviceinstances, essentially guaranteeing a concistent systemstartup and service health. That is, your NFS service will not be started before the networking service has been started up and is running. SMF is also a consistent interface to run your services/daemons with, should they be a resourcepool-daemon, a virtual-networking service, a packetfilter or a webserver - all manageable with the svcs/svcadm commands, configurable with svccfg/svcprop.  Solaris 11 utilizes SMF even heavier, you will find that the packetfilter rules can be stored in SMF, or that even the nsswitch.conf is dynamically generated at starting the relevant service.   For more information on SMF on S10 see the System Administration Guide's section about Managing Services: http://docs.oracle.com/cd/E23823_01/html/817-1985/hbrunlevels-25516.html#scrolltoc We could go on with RBAC, DTrace, JumpStart, BART and the CryptoFramework, should you be interested in those, please put your inputs in the comments below. Again: the main message is: Go for Solaris 11 if you can. If you need to run Solaris 10, we recommend deploying the mentioned technologies, they can and will improve your daily systemengineering business and prepare your platform for the move to Solaris 11. 

What do customers need the most for their UNIX platform's operation? Time. And headcount, obviously. In both of these fields there is usually room for improvement in corporations these times. That is,...

Zones? Clusters? Clustering zones? Zoneclusters?

Everyone values zones, Solaris' builtin OS-virtualization. They are near-footprintless. Their administration is delegable. They have their own bootenvironments. Easily cloneable with ZFS snapshots, etc. They are also cleanly integratable with Solaris Cluster in different ways - this post should shed some light on the different options, and provide an example of zoneclusters.  In this post I will: I. Explain the two ways to create HA services with zones on Solaris Cluster II. Show you a quick walkthrough of setting up zoneclustersIII. Envision a large platform with LDOMs, zoneclusters, central monitoring.  I. Part one, the two ways to integrate zones with Solaris Cluster Solaris Cluster has two ways to integrate Zones in a HA environment:   Failover Zones: In this case we talk about a single zone, that is monitored by the cluster, and in case of switching the service between nodes, the zone itself will be halted, detached from the global zone, moved over to another physical clusternode, attached to the Solaris instance on the global zone there, and started up.  This is a failry simple and straightforward setup, simply putting your zones to shared storage, containing the application and the data, application startup and shutdown is handled by zonestartup/shutdown.   On the other hand, the moving zone itself is a SPOF (single point of failure). If it has any issues booting, the service is down. If you do a rolling upgrade of your cluster, upgrade node A, failover and attach the zone from node B, the service in the zone is down while the upgrade-on-attach procedure is running, and should it not work out for some reason, your fallback is endangered, since the zoneupgrade has been started. Any rollback and fallback mechanism to the original global zone extends your maintenance window and hence, your downtime.  Zoneclusters: The easiest way to define zoneclusters is to compare them to zones, that they are built on. A global zone is the same to a non-global zone, like a global cluster is to a zonecluster. A global cluster is the cluster you always have installed in the global zone, having global zones as clusternodes. A zonecluster us a zone-level cluster, and has zones as clusternodes. Every zonecluster is running in a global cluster, and you can have several independent zoneclusters on a global cluster. Zoneclusters have their own cluster stack: within them you can create resources and resource groups, add or remove nodes, evacuate all nodes, shutdown the cluster, etc. Without affecting the other zoneclusters, or the global cluster. See? Just like zones.  Having two (or more) static (non-moving) zones as clusternodes has its advantages. Your application failover happens from zone to zone, removing the complexity of upgrade-on-attach and eliminating a single zone as a SPOF. Your maintenance windows are more plannable, during an update your service is running, and - should anything go wrong - you can fallback your service to a zonenode that hasn't yet been altered.  Of course, complexity doesn't simply disappear, it just changes form: You have now several clusters to manage. The advantage is, you can delegate the administration of the dozens of zoneclusters to your customers, just like you can delegate the administration of zones to the zoneadmins. This isn't anymore just a cloud solution, but a cloud without SPOF, with self service and built-in HA.  II. Part two, presenting and setting up a zonecluster I will assume that you have already an installed, initialized, running global cluster. I actually have considered including here the global cluster setup procedure, but that really is beyond the scope of this post - should a clustersetup post help you to get with clustering started, let us know in the comments, we can create a separate post. For this demo, I run Solaris 11 in VirtualBox instances, the whole magic runs on my laptop - that is, it could run on yours too. So, you have your global cluster running, and want to start with zoneclusters. What you will need is the clzonecluster command to create one. My cluster looks like this:  kvegh@sc4A:~$ clinfo && clnode list sc4Bsc4Akvegh@sc4A:~$ clresourcegroup list kvegh@sc4A:~$ kvegh@sc4A:~$ zoneadm list -cv   ID NAME             STATUS     PATH                      BRAND    IP      0 global           running    /                         solaris  shared   1 zc2              running    /zones/zc2                solaris  shared   2 zc1              running    /zones/zc1                solaris  sharedkvegh@sc4A:~$  ...that is, I have a two-node global cluster, consisting of the nodes sc4A and sc4B (Solaris Cluster 4, node A and B), no resourcegroups in the global cluster, but with two non-global zones. On node B I have a very similar zone list:  kvegh@sc4B:~$ zoneadm list -cv   ID NAME       STATUS     PATH                      BRAND    IP       0 global           running    /                         solaris  shared   1 zc1              running    /zones/zc1                solaris  shared   2 zc2              running    /zones/zc2                solaris  sharedkvegh@sc4B:~$  As you already probably assume, zc1 and zc2 are two zoneclusters, each having a zone (that is, a node) on both nodes of the global cluster. And you're right:  kvegh@sc4A:~$ clzonecluster listzc1zc2kvegh@sc4A:~$ clzonecluster status === Zone Clusters ===--- Zone Cluster Status ---Name    Node Name   Zone HostName    Status    Zone Status----    ---------   -------------    ------   -----------zc1     sc4B        sc4B-zc1         Online    Running        sc4A        sc4A-zc1         Online    Runningzc2     sc4A        sc4A-zc2         Online    Running        sc4B        sc4B-zc2         Online    Runningkvegh@sc4A:~$  Now, I told you that I have no clusterresourcegroups defined in the global cluster, but I do have one configured in the zonecluster "zc2":  kvegh@sc4A:~$ sudo zlogin zc2 [Connected to zone 'zc2' pts/2]Oracle Corporation      SunOS 5.11      11.0    December 2011You have mail.root@sc4A-zc2:~# zonenamezc2root@sc4A-zc2:~# clzonecluster listzc2root@sc4A-zc2:~# clrg status === Cluster Resource Groups ===Group Name    Node Name    Suspended   Status----------    ---------    ---------   ------apache_rg     sc4A-zc2     No          Online              sc4B-zc2     No          Offlineroot@sc4A-zc2:~#  To sum it up: On the global two-node (sc4A and sc4B) cluster there are two zoneclusters defined, zc1 and zc2. In the global zone sc4A,  within the zone zc2, that is a node in the zonecluster zc2, there is an apache-resourcegroup running that can be readily switched with usual cluster commands to the other node of the zonecluster, running on the other node of the global cluster.  On the other configured zonecluster, zc1 there are no resources configured at all, and its administrator could do all the usual cluster actions without disturbing zc2, like "clnode evacuate" or "cluster shutdown".  Now, let's configure and install a zonecluster. This is very similar to configuring and installing zones, with the only difference that the setting up of the zones will be done by the cluster:  kvegh@sc4A:~$ pfexec clzc configure newZCnewZC: No such zone cluster configuredUse 'create' to begin configuring a new zone cluster.clzc:newZC> createclzc:newZC> add nodeclzc:newZC:node> set physical-host=sc4Aclzc:newZC:node> set hostname=sc4A-newZCclzc:newZC:node> add netclzc:newZC:node:net> set physical=sc_ipmp0clzc:newZC:node:net> set address=192.168.56.163/24clzc:newZC:node:net> endclzc:newZC:node> endclzc:newZC> add nodeclzc:newZC:node> set physical-host=sc4Bclzc:newZC:node> set hostname=sc4B-newZCclzc:newZC:node> add netclzc:newZC:node:net> set physical=sc_ipmp0clzc:newZC:node:net> set address=192.168.56.183/24clzc:newZC:node:net> endclzc:newZC:node> endclzc:newZC> set zonepath=/zones/newZCclzc:newZC> verifyclzc:newZC> exitkvegh@sc4A:~$ Having configured the zonecluster, now I could start the installation of its zone-nodes, but oops! For some reason I have no access to an IPS repo. That isn't so great, since in S11 zones are not anymore created from the global zone's packages, but installed from the pkg repository directly. An alternative is: if you have a zonecluster already installed, you can clone the new ones from that one over ZFS clones, just like you can clone zones:  kvegh@sc4A:~$ clzc halt zc1 Waiting for zone halt commands to complete on all the nodes of the zone cluster "zc1"...kvegh@sc4A:~$ clzc status === Zone Clusters ===--- Zone Cluster Status ---Name     Node Name    Zone HostName   Status    Zone Status----     ---------    -------------   ------    -----------zc1      sc4B         sc4B-zc1        Offline   Installed         sc4A         sc4A-zc1        Offline   Installedzc2      sc4A         sc4A-zc2        Online    Running         sc4B         sc4B-zc2        Online    RunningnewZC    sc4A         sc4A-newZC      Offline   Configured         sc4B         sc4B-newZC      Offline   Configuredkvegh@sc4A:~$ /usr/cluster/bin/clzc clone -Z newZC -v zc1 Waiting for zone clone commands to complete on all the nodes of the zone cluster "newZC"...kvegh@sc4A:~$kvegh@sc4A:~$ clzc status newZC=== Zone Clusters ===--- Zone Cluster Status ---Name     Node Name    Zone HostName   Status    Zone Status----     ---------    -------------   ------    -----------newZC    sc4A         sc4A-newZC      Offline   Installed         sc4B         sc4B-newZC      Offline   Installedkvegh@sc4A:~$  Of course the zones have been installed too, with the name of the zonecluster:  kvegh@sc4A:~$ zoneadm list -cv   ID NAME             STATUS     PATH                      BRAND    IP      0 global           running    /                         solaris  shared   1 zc2              running    /zones/zc2                solaris  shared   2 zc1              running    /zones/zc1                solaris  shared   3 newZC            installed /zones/newZC              solaris  sharedkvegh@sc4A:~$ At this point, after starting up the newly created zonecluster newZC, you will have to configure the OS in those zones, either with "zlogin -C newZC" on both nodes, or having prepared a system profile file that you can feed the zones for automatic self-configuration, just like at any zone installation. After that you can zlogin to your zonecluster nodes, and create your HA services, manage clusterresources and resourcegroups, switch them between the zonenodes, create affinities, restart the whole zonecluster, just like with any global cluster you are used to use - and without affecting the other zoneclusters.  III. Part three, the vision of a great platform Now lean back, and imagine a platform:  This platform shall consist of 4+1 T4 servers, using LDOMs, each with two I/O Service domains that cooperate with multipathing, and two guest domains, one running Solaris 10 with zoneclusters and one running Solaris 11 with zoneclusters.  On the 4 boxes you create a 4-node S10 global cluster in the S10 ldoms, and a 4-node S11 global cluster in the S11 ldoms, deploying your applications to the platform of their choice.   You create 2-node zoneclusters within your 4-node global clusters. One 2-node ZC is for scalable applications (Webserver, DB RAC, loadbalanced Application servers, etc.), and another 2-node ZC is for failover applications (running in active-standby mode).  Having defined a role to 4 of 5 servers you decide that the 5th node should be a general HW-standby-node, to failover services to in case one of the other 4 active servers would experience HW issues.  Now to manage this platform, you can use Enterprise Manager Ops Center, to deploy/patch OS, manage virtualization, overview your clusters, monitor utilization.  Leverage the T4 capabilities, encrypt everything in HW, your ZFS, your database data, outsource your SSL encryption. Use Solaris Cluster to livemigrate your LDOMs, if you please.  With all this SPARC and Solaris 11 goodness (encryption, ZFS, virtual networking, resource management, bootenvironments), combined with zoneclusters you can build true cloud platforms, with builtin HA, self-service, application separation and high server utilization. Not to mention the sky-high coolness factor of a platform like this :)  Documentation I have used:  Oracle Solaris Cluster Software Installation GuideThe Solaris Cluster 4 Documentation Collection Footnotes:- In Solaris 10 with Solaris Cluster 3.3 there was a third option to integrate zones with the cluster, but in SC4 this isn't supported anymore, hence I did not mention it. - Solaris Cluster runs both on sparc and x86- Solaris Cluster 3.x supports Solaris 10, Solaris Cluster 4.0 runs on Solaris 11- You can run  zoneclusters with S10 and SC3.x too. 

Everyone values zones, Solaris' builtin OS-virtualization. They are near-footprintless. Their administration is delegable. They have their own bootenvironments. Easily cloneable with ZFS...

Shall I use Zones or LDOMs?

Recently (especially since the SPARC T4 release) I got this question a couple of times - "We are running/migrating to T2/T3/T4 servers, and considering for our setup the virtualization possibilities. What shall we go for, zones or ldoms?" Of course one can't answer this question without talking about the platform requirements and the reasons to pick the right technologies, but before we'd go into details, let me get the most important statement straight:Zones and LDOMs are not rivalling, but complementary technologies. If you need kernelspace separation, use ldoms. But run your applications in zones within those ldoms anyway!Let's get some terminology clear first: LDOMs are now called Oracle VM for SPARC. I will use these terms interchangably.  Zones have started their lives as project Kevlar, then named zones, then marketed as containers, we are now back to zones again.   LDOMs are the HW-Virtualization technology of the SPARC-T (CMT, ChipMultiThreading, Coolthread, sun4v, etc) server series, it is their ability to carve up the server into Logical DOMains, running on a hypervisor that runs in the firmware.  Zones are the featherweight OS-Virtualization technology of Solaris on all of the platforms (Sparc-T, Sparc-M and x86 too) Every T server is running ldoms. If you don't partition your box into domains, you are still running one single large ldom, called the primary domain, encapsulating the complete server.  Every Solaris 10+ OS installation has one zone, the global zone (GZ). This is where the [shared] kernel[space] runs, and the non-global zones (NGZ) are the containers separating applications in the userspace.  Now, why would you want to run zones?  Container principle: They cleanly separate your applications from each other, by maintaining for them a separate set of Solaris packages, their dedicated CPU resources, their IP-stack, their filesystems, etc.  Clean architecture: You won't poison your OS installation in the GZ running on the HW with additional packages/settings. The GZ manages the resources between the zones, runs the kernel, does the scheduling, runs the cluster, manages the devices, etc. The NGZs run the applications. Flexibility: You can simply detach a zone from the GZ and attach it to another GZ on another box, including the application. You can easily clone zones too.  Security: Should a NGZ ever get compromised, the attacker can't bother the GZ, or applications running  in other NGZs. Resource Management: You can dedicate the guaranteed amount of CPU shares a zone should get (using the FairShareScheduler), but as long as your CPU pool isn't 100% utilized, every zone can use more than the amount dedicated to it - that is, you can overcommit your resources.  And what are the reasons to run LDOMs?  Kernel level separation:  You might want to run different updates of Solaris 10 within a box.  You might want to run Solaris 10 and Solaris 11 right next to each other within a box.  Live migration: You can't livemigrate zones, but you can livemigrate ldoms.  Some of your applications might require to run in the GZ, and you don't like the idea of running applications both in the GZ and its NGZ at the same time, hence you separate them into ldoms.  You need to reduce the number of vCPUs in a box for licensing issues. LDOMs are now recognized as hardpartitions by Oracle, license boundaries.  You don't want your I/O to depend on a single service domain - you can build multipathgroups of devices between two I/O device providing service domains.  As you see these two technologies fulfill different requirements, they are in different levels of your operation-stack, ldoms being a HW-virtualization - a host for kernels to run, and zones being an OS-virtualization, to provide containers for your application to run in:   to give you an idea: run S10 and S11 in ldoms next to eachother within the same box, run branded and native zones on top of them To summarize: The question shouldn't be about zones vs. ldoms. Use zones, they are your friends. The question is, if you partition your T-SPARC server into ldoms below your global zones to run your NGZs in. Especially with Solaris 11, with Crossbow, the new network virtualization technology (that enables all your NGZs to have a dedicated IP stack) and the possibility to run Solaris 11 native zones and Solaris10 branded zones on top of Solaris 11, you have two quite powerful technologies to really get your server's worth - and by that I mean having a high server utilization. The higher that utilization is, the more you get for your costs.  Additional Information: LDOMs WikiOracle VM for SPARC DocumentationWhite Paper: Best Practices For Network Availability with OVM for SPARCWhite Paper: Best Practices For Data Availability with OVM for SPARCHOWTO: Getting started with Solaris ZonesOracle Solaris Zones Documentation Best Practices running Oracle DB in Zones Running RAC in ZoneClusters

Recently (especially since the SPARC T4 release) I got this question a couple of times - "We are running/migrating to T2/T3/T4 servers, and considering for our setup the virtualization possibilities....

Social Network Sites, Blog recommendations for Oracle Systems

Blogs, social networks, discussion groups collect, converge and forward uptodate news and information about every field one can imagine - IT, Hardware, virtualization and development are prime examples to this rule. To ease the hunt for the sources, we have collected a set of links in some of the most popular social networks of your choice and included some blog links for your rss reader too. Feel free to add them to your Google+ circles, follow them on Twitter and 'Like' them on Facebook, and add the blogs to your feed.   on G+ Oracle Solaris Oracle Hardware Oracle University Solaris Training Oracle Infrared Oracle Partner Network The central Oracle G+ site on Twitter  Oracle Solaris Oracle Hardware Oracle Virtualization Solaris Studio Oracle Blogs Oracle University The central Oracle Twitter handle Solaris Certified Oracle Enterprise Manager Oracle Partner Network OTN Garage My Oracle Support OTN ArchBeat Oracle Engineered Systems on Facebook Solaris Solaris Certified Oracle Technology NetworkOracle Enterprise Manager Ops Center Oracle HW Oracle Virtualization Exalogic Elastic Cloud Solaris Studio OTN Garage Oracle Partner Network OTNArchBeat OracleSunJP: Oracle HW Systems information in Japanese The central Oracle FB site onBlogs of course the Oracle Systems Blog Austria :) Oracle Solaris Blog OTN GarageThe Oracle Virtualization BlogOracle Infrared blog The Oracle Enterprise Manager blog Stefan Hinker's Sparc/Solaris blog Darren Moffat's blog Tim Cook's blogPerformance and Best PracticesIsaac Rozenfeld's blog Dan Anderson's blog Gerry Haskin's patch blogSolaris TipsSolaris and System Information for ISVsGerman speaking Solaris blogGerman-speaking PartnerblogOpenomicsGiri Mandalika's blogChris Weal's blogJean-Christophe Lamoure's ClusterblogZFS Storage Appliance TipsPaul Johnson's blog OTN ArchBeat blogAnd generally: http://blogs.oracle.com - browse by topics/tags!  Enjoy!  The Hardware Presales Consultant Team from Austria 

Blogs, social networks, discussion groups collect, converge and forward uptodate news and information about every field one can imagine - IT, Hardware, virtualization and development are...

Building an Infrastructure Cloud with Oracle VM for x86 + Enterprise Manager 12c

Cloud Computing? Everyone is talking about Cloud these days. Everyone is explaining how the cloud will help you to bring your service up and running very fast, secure and with little effort. You can find these kinds of presentations at almost every event around the globe. But what is really behind all this stuff? Is it really so simple? And the answer is: Yes it is! With the Oracle SW Stack it is! In this post, I will try to bring this down to earth, demonstrating how easy it could be to build a cloud infrastructure with Oracle's solution for cloud computing.But let me cover some basics first: How fast can you build a cloud?How elastic is your cloud so you can provide new services on demand? How much effort does it take to monitor and operate your Cloud Infrastructure in order to meet your SLAs?How easy is it to chargeback for your services provided? These are the critical success factors of Cloud Computing. And Oracle has an answer to all those questions. By using Oracle VM for X86 in combination with Enterprise Manager 12c you can build and control your cloud environment very fast and easy. What are the fundamental building blocks for your cloud? Oracle Cloud Building Blocks #1 Hardware Surprise, surprise. Even the cloud needs to run somewhere, hence you will need hardware. This HW normally consists of servers, storage and networking. But Oracles goes beyond that. There are Optimized Solutions available for your cloud infrastructure. This is a cookbook to build your HW cloud platform. For example, building your cloud infrastructure with blades and our network infrastructure will reduce complexity in your datacenter (Blades with switch network modules, splitter cables to reduce the amount of cables, TOR (Top Of the Rack) switches which are building the interface to your infrastructure environment. Reducing complexity even in the cabling will help you to manage your environment more efficient and with less risk. Of course, our engineered systems fit into the cloud perfectly too. Although they are considered as a PaaS themselves, having the database SW (for Exadata) and the application development environment (for Exalogic) already deployed on them, in general they are ideal systems to enable you building your own cloud and PaaS infrastructure. #2 Virtualization The next missing link in the cloud setup is virtualization. For me personally, it's one of the most hidden "secret", that oracle can provide you with a complete virtualization stack in terms of a hypervisor on both architectures: X86 and Sparc CPUs. There is Oracle VM for X86 and Oracle VM for Sparc available at no additional  license costs if your are running this virtualization stack on top of Oracle HW (and with Oracle Premier Support for HW). This completes the virtualization portfolio together with Solaris Zones introduced already with Solaris 10 a few years ago. Let me explain how Oracle VM for X86 works: Oracle VM for x86 consists of two main parts: - The Oracle VM Server: Oracle VM Server is installed on bare metal and it is the hypervisor which is able to run virtual machines. It has a very small footprint. The ISO-Image of Oracle VM Server is only 200MB large. It is very small but efficient. You can install a OVM-Server in less than 5 mins by booting the Server with the ISO-Image assigned and providing the necessary configuration parameters (like installing an Linux distribution). After the installation, the OVM-Server is ready to use. That's all. - The Oracle VM-Manager: OVM-Manager is the central management tool where you can control your OVM-Servers. OVM-Manager provides the graphical user interface, which is an Application Development Framework (ADF) application, with a familiar web-browser based interface, to manage Oracle VM Servers, virtual machines, and resources. The Oracle VM Manager has the following capabilities: Create virtual machines Create server pools Power on and off virtual machines Manage networks and storage Import virtual machines, ISO files, and templates Manage high availability of Oracle VM Servers, server pools, and virtual machines Perform live migration of virtual machines I want to highlight one of the goodies which you can use if you are running Oracle VM for X86: Preconfigured, downloadable Virtual Machine Templates form edelivery With these templates, you can download completely preconfigured Virtual Machines in your environment, boot them up, configure them at first time boot and use it. There are templates for almost all Oracle SW and Applications (like Fusion Middleware, Database, Siebel, etc.) available. #3) Cloud Management The management of your cloud infrastructure is key. This is a day-to-day job. Acquiring HW, installing a virtualization layer on top of it is done just at the beginning and if you want to expand your infrastructure. But managing your cloud, keeping it up and running, deploying new services, changing your chargeback model, etc, these are the daily jobs. These jobs must be simple, secure and easy to manage. The Enterprise Manager 12c Cloud provides this functionality from one management cockpit. Enterprise Manager 12c uses Oracle VM Manager to control OVM Serverpools. Once you registered your OVM-Managers in Enterprise Manager, then you are able to setup your cloud infrastructure and manage everything from Enterprise Manager. What you need to do in EM12c is: ">Register your OVM Manager in Enterprise ManagerAfter Registering your OVM Manager, all the functionality of Oracle VM for X86 is also available in Enterprise Manager. Enterprise Manager works as a "Manger" of the Manager. You can register as many OVM-Managers you want and control your complete virtualization environment Create Roles and Users for your Self Service Portal in Enterprise ManagerWith this step you allow users to logon on the Enterprise Manager Self Service Portal. Users can request Virtual Machines in this portal. Setup the Cloud InfrastructureSetup the Quotas for your self service users. How many VMs can they request? How much of your resources ( cpu, memory, storage, network, etc. etc.)? Which SW components (templates, assemblys) can your self service users request? In this step, you basically set up the complete cloud infrastructure. Setup ChargebackOnce your cloud is set up, you need to configure your chargeback mechanism. The Enterprise Manager collects the resources metrics, which are used in a very deep level. Almost all collected Metrics could be used in the chargeback module. You can define chargeback plans based on configurations (charge for the amount of cpu, memory, storage is assigned to a machine, or for a specific OS which is installed) or chargeback on resource consumption (% of cpu used, storage used, etc). Or you can also define a combination of configuration and consumption chargeback plans. The chargeback module is very flexible. Here is a overview of the workflow how to handle infrastructure cloud in EM: Summary As you can see, setting up an Infrastructure Cloud Service with Oracle VM for X86 and Enterprise Manager 12c is really simple. I personally configured a complete cloud environment with three X86 servers and a small JBOD san box in less than 3 hours. There is no magic in it, it is all straightforward. Of course, you have to have some experience with Oracle VM and Enterprise Manager. Experience in setting up Linux environments helps as well. I plan to publish a technical cookbook in the next few weeks. I hope you found this post useful and will see you again here on our blog. Any hints, comments are welcome!

Cloud Computing? Everyone is talking about Cloud these days. Everyone is explaining how the cloud will help you to bring your service up and running very fast, secure and with little effort. You can...

The SPARC T4 servers are here!

Ladies and gentlemen, it is an honour for us to announce: The SPARC T4 processors and the servers built around them are released. Before we'd dive into the features, a short summary of our SPARC platform:  The M-Series are designed with Mainframe-class RAS features (Reliability, Availability, Serviceability). They are based on the Sparc64-VII+ CPUs, excelling at single threaded performance. The T-Series are the CoolThread servers, with the CMT (chipmultithreading) design, they are designed to run heavily parallel workloads, concentrating on throughput, running up to 512 threads actively at the same time, if desired.  The latter category just got a brand new update, let's see, what makes the T4 special:  Just like Solaris 11 was greatly influenced by the community's requests, the T4 has been designed based on listening to customer feedback about previous (T2+, T3) processors:  "The throughput of the T3 is awesome, but for some workloads a higher single thread performance is desired"  "ChipMultiThreading with many cores times many threads is nice, but how about a higher clockfrequency?"  "Make sure to keep/enhance the System-on-a-Chip advantages, like crypto-HW, OnChip 10G NIC and PCIe" Engineering has added:  "L2 cache is great, but you know what made the UltraSparc VI+ that powerful? L3 cache." "Stick to  SPARC V9 and the well-received CMT model" "Let's move the cryptoaccelerators from coprocessors into the chip itself and drive cryptography within the pipeline for crypto performance" "Using out-of-order execution again would bring performance enhancements too"  Management requested: "Stick to high parallellity, high throughput and power management features." "Reuse successful concepts, release time is critical for this product" "If using the 28nm technology means delays, then stick to 40nm for now, and use 28nm for the next CPU." Engineering has been working for a while on the new S3 cores, that have replaced previous generetion cores. The S3 core fulfills all the requirements, and more. It has a shared 4MB L3$, and a private L2$ on each core. It has out-of-order execution, which allows it to run non-dependent instructions parallelly out of call-order, proactively. With 8 next-gen S3 cores the T4 still matches the throughput of a T3 (which had 16 cores), and still brings a 5x higher single thread performance compared to T3. (By the way - a 3x higher single thread performance was planned, but engineering excelled themselves once again). These features make the T4 to a much more general purpose processor.  Now, where can one use single threaded performance at this levels? For example on databases. Or, LDOM live migration times have been significantly improved as well with the T4. Or at any other application that aren't designed with purely parallel workload in mind. Let's see a graphical representation:  A very important feature is what we call critical thread: It is the capability of the CPU to provide access to a complete core for a running thread. Your applications throughput may be limited by the serial portions of your workload. Setting the priorities of those sequentially working threads high raises the attention of the scheduler, and it can decide to dedicate a complete core to a thread instead of running 8 threads parallelly on that core. A noteworthy feature, especially for all this happens dynamically, on-the-fly. If your application needs throughput, it gets 8 threads. If it needs single threaded performance, it gets a complete core. This is a very seemless process, without the administrator having to switch CPU-modes.  Allow us to mention some - in our opinion - very much undermarketed features of the T series:  OnChip 10G NICs. This actually means that you have two 10Gbps network interfaces sitting directly on the CPU. Data does not need to fight its way through the PCI bus labyrinth, you do not need to use additional NICs with network logic built into them, all you need is an XAUI converter to lead the onchip-nics to the outer world. This way you have very low latency.  Oracle VM for SPARC (former LDOMS): LDOMs are supported as a matter of course, customers can continue to partition these servers and run several different Solaris instances (even Solaris 10 and 11 mixed) next to eachother, separated by the hypervisor running in the firmware. Without a price tag! Powermanagement: The T4 can do cycle skipping, just like in T3, to lower power consumption.  Encryption: Solaris supports an extensive set of cryptographic algorithms, and the T4 can provide encryption services implemented in hardware, not having to software compute those. On-the-fly, no-cost, NSA-approved encryption.   An encryption performance comparison between T3 and T4:  The T4 is supported in Solaris 10 u10 (u8 and u9 with patches) and in Solaris 11. Oracle applications on the SuperCluster will be able to benefit from the critical thread capability too. We think this is actually a prime example of Hardware and Software engineered to work together, or even Hardware-, OS- and even Applications engineered to work together.  Just like Solaris 11, the T4 CPU was designed very much with supporting Oracle software in mind, providing features those can benefit from. Oracle has the very unique situation of being able to not only to tune the software elements to the requirements, but also the operating system and the hardware it is running on. This is the key differentiator of the Oracle products. Integration throughout the whole stack. Oracle has released 1-, 2- and 4 Socket servers, the T4-1, the T4-2 and the T4-4. These have been delivered as per the SPARC Roadmap, which Oracle has committed to not to deliver later than defined. This is another Oracle-specific feat - no other CPU vendor declares committment publicly 5 years in advance.  And now, the time for consolidation has arrived. One T4-2 can replace several old V-servers with UltraSparcIII or IV, not to mention the T1-T2 servers. Of course it makes also a lot of sense to bring your Oracle SW back from other HW and OS to the OS and the Systems they run best on - Oracle on Oracle. Ask your local HW presales representative for upgrade paths.

Ladies and gentlemen, it is an honour for us to announce: The SPARC T4 processors and the servers built around them are released. Before we'd dive into the features, a short summary of our...

The most inviting Solaris 11 features - Part I, BootEnvironments

Disclaimer: "Having been involved in various projects around the upcoming Solaris 11 release, we had the possibility to compile a list of features we assume UNIX Engineers will find to be cornerstones of Solaris 11 based platforms. It wasn't easy to keep the list short, due to the sheer amount of innovation and the tight integration of the new- or updated technologies in this Solaris release. We have planned a series of blog posts with a short preview of each Solaris 11 technologies in the list, in a Question and Answer form." This is the first post, featuring BootEnvironments.And here we go with the questions:  Q: What is a bootenvironment? A: It is a set of filesystems, that together build a complete, bootable Solaris instance. After a fresh Solaris 11 install, right after the first boot, the filesystems created - together with their bootable capability - form the first bootenvironment.  Q: How is this different from LiveUpgrade ABEs on Solaris 10? A: BooteEnvironments (BE) have evolved from the LiveUpgrade technology. The main differences are: BootEnvironments were integrated to be a core element in Solaris 11, and are not an add-on technology option anymore.  BootEnvironments are the main and only way of managing Solaris 11 OS upgrades.  ZFS is the choice of filesystem, BEs by default are created on ZFS clones, essentially wasting no storage space.  One BE is auto-created at installation.  IPS Integration: The new Image Packaging System provides hooks to manage BEs transparently at package installation/upgrade times. A system upgrade through 'pkg update' automatically creates a new BE in the background, upgrades the packages there, and awaits your reboot into the new BE. Single packages can request BE creation or snapshot at package installation too.  Non-global Zones contain bootenvironments too, and are associated with the parent bootenvironment in the global zone. Q: What are the main goals using this technology? A: Minimizing downtime, fallback mechanisms and delegated administration:  System upgrades can be done on alternate, cloned BootEnvironments, without interrupting productive operations, maintenance windows can basically be reduced to the reboot interval.  Should a new, modified BootEnvironment not fulfill all the expectations, you can boot any of the previously created unaltered BEs. Or to put it another way: Should anything happen to the active bootenvironment, you can any time revert to a clone you created.  Some bootenvironment management tasks (snapshot, activate, clone...) can be delegated to zone administrators, who can use the beadm command within the non-global zone, effectively managing their bootenvironment from within their zone. Now let's see an example. You get the task to prepare an environment for an SAP deployment. No problem, you login to the right host (in our case: host1), create a new bootenvironment in it, activate the BE, reboot into it, and hand it over to the SAP administrators for software setup:  Step 1: What is the current BE setup/status on the host?  root@host1:~# beadm listBE        Active Mountpoint Space  Policy Created--        ------ ---------- -----  ------ -------solaris-0 NR     /          46.59M static 2011-09-24 15:19root@host1:~# Step 2: Create the new BE:  root@host1:~# beadm create SAProot@host1:~# beadm listBE        Active Mountpoint Space  Policy Created--        ------ ---------- -----  ------ -------SAP       -      -          36.0K  static 2011-09-24 16:19solaris-0 NR     /          46.59M static 2011-09-24 15:19root@host1:~# Step 3: activate BE root@host1:~# beadm activate SAProot@host1:~# init 6root@host1:~# [...waiting for the host to come back up...] Step 4: Login, check, handover:  root@host1:~# beadm listBE        Active Mountpoint Space  Policy Created--        ------ ---------- -----  ------ -------SAP       NR     /          77.51M static 2011-09-24 16:19solaris-0 -      -          8.29M  static 2011-09-24 15:19root@host1:~#root@host1:~# echo "host1 is prepared for SAP setup" | mailx -s "SAP host" \-c unixadmins@corp.com -r unixadmins@corp.com sapadmins@corp.comroot@host1:~# Step 5: The lengthy procedure of SAP setup. sap@host1:~# ./setup_sap.sh sap@host1:~# Of course it is not this easy, but this is a bootenvironments-post :)  Step 6: After the application setup, create a snapshot of the BE with the SAP installation in it, just in case... root@host1:~# beadm create SAP@backuproot@host1:~# beadm list -aBE/Dataset/Snapshot                   Active Mountpoint Space  Policy Created-------------------                   ------ ---------- -----  ------ -------SAP   rpool/ROOT/SAP                     NR     /          66.74M static 2011-09-24 16:19   rpool/ROOT/SAP@2011-09-24-14:19:18 -      -          10.85M static 2011-09-24 16:19   rpool/ROOT/SAP@backup              -      -          0      static 2011-09-24 16:42solaris-0   rpool/ROOT/solaris-0               -      -          8.29M  static 2011-09-24 15:19root@host1:~# The next day, during the first SAP tests, an SAP admin calls you with panic in his voice, during SAP upgrade a procedure ran amok and they'd need to restart the upgrade from scratch. You calm him, explaining that your system is well prepared for cases like this, and you can revert all the changes to the point where they had their clean SAP install done, so they can start upgrading again.  Step 7: Create a BE based on the SAP@backup snapshot, activate, and boot it, destroy the old SAP BE, snapshot the new SAP2ndtry BE too, you know, just in case... root@host1:~# beadm list -sBE/Snapshot                Space  Policy Created-----------                -----  ------ -------SAP   SAP@2011-09-24-14:19:18 10.85M static 2011-09-24 16:19   SAP@backup              24.0K  static 2011-09-24 16:42solaris-0root@host1:~# beadm create -e SAP@backup SAP2ndtryroot@host1:~# beadm listBE        Active Mountpoint Space  Policy Created--        ------ ---------- -----  ------ -------SAP       NR     /          77.67M static 2011-09-24 16:19SAP2ndtry -      -          36.0K  static 2011-09-24 16:58solaris-0 -      -          8.29M  static 2011-09-24 15:19root@host1:~# beadm activate SAP2ndtryroot@host1:~# init 6root@host1:~# [...wait for the host to come up...] root@host1:~# beadm destroy SAPAre you sure you want to destroy SAP?  This action cannot be undone(y/[n]): yroot@host1:~# beadm listBE        Active Mountpoint Space  Policy Created--        ------ ---------- -----  ------ -------SAP2ndtry NR     /          86.90M static 2011-09-24 16:58solaris-0 -      -          8.29M  static 2011-09-24 15:19root@host1:~# beadm create SAP2ndtry@backuproot@host1:~# You have saved their day, and a lot of effort of completely having to re-setup their application from scratch. Even if they'd decide to go for PeopleSoft instead of SAP, all you'd need to do is to create a BE cloned from the original solaris-0 BE, and you have a completely clean environment again for the PeopleSoft tests.   And now, the surprise: All these reboots were not harmful for any other non-global zones, because...:  root@host1:~# zonenamehost1root@host1:~# exitlogout[Connection to zone 'host1' pts/2 closed]root@s11ea:~# zoneadm list -cv  ID NAME             STATUS     PATH                           BRAND    IP   0 global           running    /                              solaris  shared   7 host1            running    /zones/host1                   solaris  excl   - testz            installed  /zones/testz                   solaris  exclroot@s11ea:~# ...because all this has happened within a zone! We have managed the BE in a zone, no other zones were harmed in the making of this blogpost.  To summarize: Having had separated bootenvironments allowed me to separate different configurations and application deployments from eachother, easily being able to revert to a desired snapshotted state. In this case, we have operated on the bootenvironments within a zone, haven't touched the bootenvironments of the global zone.  Another example could be upgrading the operating system in a freshly cloned BE, rebooting there, and if something would go wrong with the new release - simply reactivate the old BE, and use the old release of the OS. This, of course brings up the question: Q: What if I upgrade the OS, but it goes terribly wrong, and it can't even boot? How do I revert? A: Generally, Solaris is very well tested before a release is published. But should the case still happen, beadm auto-updates the bootmenus: You can choose to boot different BEs from the grub menu at boottime on x86, or choose the BE from OBP using the 'boot -L' option on sparc.  We could go on and on for possible usecase scenarios, allow me to stop here and provide the links for further information:  Q: Where can I find the documentation about bootenvironments? A: Solaris 11 has not been released yet, but you can download the Solaris 11 Early Adopters' release, the site also has a link to the downloadable documentation package. 

Disclaimer: "Having been involved in various projects around the upcoming Solaris 11 release, we had the possibility to compile a list of features we assume UNIX Engineers will find to be cornerstones...

Solaris 10 u10 has been released. What's new?

Solaris 10 u10 (8/11) has been released, including documentation, patches, DVD images, the whole package.  There are numerous featureupdates in performance-, networking-, security-, administration-, and HW-support-fields, let's see some appealing examples:  ZFS diff: It is well established, that ZFS is a Copy-on-Write filesystem, that is, if you snapshot/clone a filesystem, only the modified blocks will take up storage. But how do you know what exactly has changed between two snapshots among the files? That is where the 'zfs diff' subcommand comes handy, enlisting differences (files created/removed/modified).  SPARC: Support for fast rebootRepetitively having to wait for a larger server to reboot can be timeconsuming, because of the hardware testing (POST). Now you can skip certain tests, with the reboot -f option. You also have the possibility to set this behaviour as default for every reboot. The only question that remains to the latter is if you would want to run on not thoroughly tested hardware? One doesn't reboot a Solaris server that often. This feature is more about minimizing downtime. New tools to analyze CMT loads: pginfo and pgstat With these two new tools one can display CPU structures and break down the more complex utilization of the ChipMultiThreading processors (SPARC-T Series). For examples see two blogposts: by Darryl Gove: http://blogs.oracle.com/d/entry/pginfo_pgstatand by Stefan Hinker: http://blogs.oracle.com/cmt/entry/solaris_knows_hardware_pgstat_explains Oracle Configuration Manager: This agent will enhance supportability by collecting information about your server, optionally uploading it to the Oracle Support improving reaction times and supportability. Yes, it can operate in disconnected mode.  Performance Improvements:Just to pick two of the many: libmtmalloc, the multithreaded memory allocator library has been revised to provide better performance for 64bit applications with a large number of threads. That is, Java applications can gain a performance boost. Also [Dynamic] Intimate Shared Memory actions have been improved that will affect certain Oracle database operations. Now, why would I mention exactly these two over the others? Because they make a point of how Oracle attaches importance to integration, effectively supporting database and application performance from the operating system level. Just like our own HW (T4, anyone?) is built to work tightly together with Solaris, that is supporting the application layer, to what our applications rely on.  Security Enhancements: For one, the Oracle Key Manager now cooperates with the Solaris CryptoFramework. (Have I mentioned integration already?) Secondly, the SSH daemon supports now chrooting users to directories. Can come handy if you are hosting on a lower level where zones would be an overkill.  The "What's new" document reveals numerous additional new features about the Oracle Configuration Manager, VTS improvements to selectively test HW components (I/O buses, CPUs, disk performance...), the extended Intel SandyBridge instruction-set support, and so forth, see: http://download.oracle.com/docs/cd/E23823_01/html/821-2730/gijtg.html#scrolltoc ).  What about support for the T4? Well, the "What's New" doc doesn't explicitly mention it, but if you look into the README of kernelpatch 144500-12 bundled with u10, you will find the mos welcome message stating "Solaris support for SPARC T4 platforms". That is right, Solaris 10u10 supports the impatiently anticipated SPARC T4.  This is another important message. With the very-near-future release of Solaris 11 Solaris 10 will not be dropped. It just got an update to a platform that hasn't even been released yet. Of course, the features of Solaris 11 are very inviting. Upgrade if you can. But we also understand the safe-harbour-policy, never-touch-a-running-system corporate culture. See the Solaris roadmap, there will be more Solaris 10 updates. Or, see the Support Policy for HW and Operating Systems, Solaris 10 has premier support right up to 2015!  Solaris 11 is the future. But rest assured, that we stick to Solaris 10 too. 

Solaris 10 u10 (8/11) has been released, including documentation, patches, DVD images, the whole package.  There are numerous featureupdates in performance-, networking-, security-, administration-,...

Introducing the Hardware Presales Consultant Team in Austria

Welcome to the brand-new blog of the Oracle Hardware Presales Team in Vienna!  Allow us to introduce ourselves, and our intentions with this newly started web-presence on Oracle Blogs.   We are a team of four consultants:   Alexander Kern is the Teamlead of the local Hardware Presales Team. He has joined Sun Microsystems in 1999 in Munich and works in Austria since 2005. During his 12 year work in IT industry in different roles from System Engineer, Engagement Manager, Project Manager and Team Leader he developed a broad knowledge about almost all kind of datacenter infrastructure.  Richard Rotter is active since 1985 in the IT market, joined Sun Microsystems in 2001 with Oracle DBA background.  His main focus at Oracle is HW Server business, e.g. SPARC and x86 servers, the virtualization technology OVM for x86 and conducting Exadata Proof of Concepts.  Harald Schmaldienst joined Oracle in 2011, after years of being a UNIX/Linux system administrator, IT Architect, DBA Teamleader and finally Presales Consultant in Austria's financial sector. Besides being a definitive SPARC supporter Harald is convinced that ORACLE's Optimized Solutions and Engineered Systems definately make the difference for today's datacenter and SPARC Supercluster will take over the world. Karoly Vegh has joined Oracle in 2011. He has worked as a UNIX System Engineer maintaining Solaris and Linux servers for over a decade before doing so at ISPs, banks and online entertainment companies. He is a great supporter of the Solaris operating system, and believes that with the Solaris Zonecluster and the EM Ops Center management one can build truly consolidated platforms. Our Team is in charge in Austria to inform customers and provide solutions on the technology level around the complete Oracle Hardware portfolio, including: SPARC (both M and T series) and x86 servers Operating Systems Oracle Solaris and Oracle Linux Virtualization technologies OVM for SPARC (former LDOMs) and OVM for x86, Zones and Dynamic Domains In the storage field NAS, SAN (and SAM!) to backup and archiving projects High Availability with Oracle Solaris Cluster Datacenter Management with Oracle EM Ops Center The engineered solutions Exadata Database Machine, Exalogic and SPARC Supercluster Serverconsolidation projects, platform design recommendations, HW TCO calculations Technology-update presentations for both partners and customers ...and many adjacent fields. These are the topics we plan to blog about here as well. Should we be able to support your projects, provide information about possible solutions around the Oracle HW products - do not hesitate to contact us! 

Welcome to the brand-new blog of the Oracle Hardware Presales Team in Vienna!  Allow us to introduce ourselves, and our intentions with this newly started web-presence on Oracle Blogs.   We are a team...