Wednesday Nov 04, 2009

A Time Capsule for the Enterprise?

On Halloween I upgraded the desk side PC from Windows XP to Windows 7.  Between multiple reboots, application installs, recovery of email, print server configuration, etc. my wife asked "Why does it have to be so hard?"  As she said this I was looking at the Apple Time Capsule sitting on top of a small cabinet in our home basement. The only item coming out of the Time Capsule was the power cord connected to the power outlet.  I run the Time Capsule as a wireless client in our home network for data backup.  The data consists of many pictures, video clips, songs, the kids homework and basically many various files (some of which are important).

My home (LAN) network has grown over the years from a few PCs connected via an old 802.11b router to a dual band (802.11g/802.11n) router connecting a multitude of wireless clients. These include a XBox, iTouch, PrintServer, OpenSolaris, Ubuntu, Mac, Windows XP, Windows Vista and Windows 7 devices.  Basically a heterogeneous environment.  Will the older Window versions be phased out at home eventually?  Yes.  The Unix/Linux clients will remain out of necessity as well as for hobby.  It's difficult to currently beat the Mac experience.  Adding a Time capsule into my existing home network was relatively easy.  The Airport setup updated the Time Capsule firmware and configured the device in straight forward steps.  It wasn't as "fun" for a "techie" as say CLI commands, but simple is defined as "fun" for most consumers. The Windows XP to Windows 7 full upgrade was painfull but I have to admit the network configuration experience was much improved.  I was pleasantly surprised freeware Bonjour discovery services just worked and the W7 system configured the Time Capsule as a usable share.

One could imply a similar situation in the enterprise space.  For an enterprise business "fun" is defined as high margin dollars on a growing revenue stream.  This usually means your costs are contained, you implement continuous improvements on efficiency and you simplify.  Vendors in the technology industry are all trying to provide a truly "Enterprise Time Capsule" or Appliance.  There has been a large amount of innovation over the past 20 years.  However today's innovation is tomorrow's PDP-11Minicomputers were appliances that Mainframes couldn't be.  The evolution has continued over the years in every technology segment.

Enterprise customers want it simple as do consumers.  While the stakes are much higher in the enterprise, the bar is significantly raised for "just working" each and every time.  It doesn't matter if you are playing catchup, you are the incumbent or you are the new thought leader-- the winners will be the set of vendors who provide the tool that "just works" each and every time in the harshest and most complex environments. 

Blog is available also at: http://bobporras.wordpress.com/


Thursday Mar 20, 2008

A Completed Open Source Storage Stack... no kidding

Another storage code base has been posted at opensolaris.  This most recent contribution focuses on the area of hierarchical storage management (HSM).  The technology is much more than standard backup.  It addresses automated data management via policies driven by data and metadata.  HSM drives some of the largest data repositories out there in the industry today.  With our open sourcing of SAM-Q we have completed an extremely large complex effort of open sourcing our \*entire\* storage stack!

This milestone is only the end of the beginning since we have many new open storage projects in process at opensolaris.org. These new storage projects are all being developed out in the open with the community.  Everything from data services, protocols, file systems, compression, encryption, replication, snapshots, drivers and archive software is available to the community.  There is no other comprehensive open sourced storage stack out there in the industry.

However, there are other comprehensive proprietary storage stacks out there that are quite good but you pay a hefty price (premium) for each part of the storage stack.  If you have the time but not the money the opensolaris community may be the place where you can contribute.  The community may also be the place for you if you are trying to establish your business or solution at a revolutionary price point.  In either case pure economics is a driving force. 

You may want to check out my most current read. "Alan Greenspan - The Age of Turbulence"  If you enjoy economics, history and want to ponder the power of the open source movement (aggregate demand ;-) )-- this book is a must read.  Alan also gives you some insight into the current market meltdown.

Read what some of the many team members Margaret Hamburger, Ted Pogue and Lynn Rohrer have to say about our latest opensource efforts. The entire team's pace and execution responded to a very aggressive goal set by me ~1 year ago... "Open Source the entire storage stack."  I'm also excited by the code contributions made by partners and vendors to the community.  It is also equally exciting to see customers using the open sourced storage technologies to build their storage products for their businesses. 

We at Sun also have the opportunity to build hybrid storage solutions with the opensolaris storage stack as well.  After all, open source software is in our DNA and we are the largest commercial contributors of open source software in the world.  A big thank you to the entire team.


Thursday Feb 28, 2008

Green Eggs and XAM

I am XAMXAM I am.  Sun recently made another open source contribution-- this time in the space of fixed content storage software.  Fixed content storage is growing at an exponential rate.  As an example, stop and imagine how much fixed content data will be generated by the Summer Olympics 2008. Pictures, video, tickets, security data, statistics, invoices, hotel bills, airline reservations, etc, etc. etc.  A massive amount of data that will be stored and preserved digitally for a long time.  Customers depend on Archive Products, even better if they are open.

There is now a fully open source code base contributed to opensolaris.org, java.net and the SNIA XAM TWG.  Read Scott Tracy's perspective. This code base eliminates the need to roll your own digital archive using piece parts (server, RAID HW, database, etc.)  Yet another storage solution built on the equation:

Powerful considering this implementation is not proprietary and uses industry standard APIs. Our contribution to XAM further emphasizes the commitment to eliminate the barriers in the fixed content storage arena. No more closed APIs to a specific vendors hardware or software stack, but rather as an industry standard such as Ethernet. Some vendors are being forced to open their APIs as opensource is having a positive effect for customers.

Code contrbutions can be found at java.net, opensolaris.org and for more info on joing SNIA please go here.

A great start for building your own digital archive appliance with proven enterprise software that is available as an open source code base.  A common theme for some vendors that are leading the way


Thursday Jan 17, 2008

Don't Forget That Storage Plumbing

Few can say that storage is not one very hot topic in the IT sector today.  The internet has created quite a medium for content delivery of podcasts, blogs, webcasts, webinars, etc.  This blog itself is an example.  Storage solutions are abundant, but don't forget that some storage vendors charge you for every neat feature.  Yes you pay for that special hardware that does compression to save storage space.  You also pay for every one of those protocols that you need (iSCSI, CIFS, NFS, etc.) and let's not forget about those important data services for protection (replication, clustering, anti-virus engine, etc.)  Some vendors even charge big money when you grow out of your storage pool and want expansion.  Can you say fork lift upgrade?  Well not all storage vendors want to charge you twice.  As the storage market approaches commodity some of us are getting ahead of the curve.  It is true that all storage companies want to make money... but the difference is leading the way versus fighting against something that will happen anyway.  Even new fast storage hardware will become a commodity as others join with similar offerings. It is important to understand that storage today is heavily reliant on low level software such as device drivers, frameworks and protocols that enable the higher level software in the storage stack to simply work.  If a company can expose and open up the storage stack it has a good possibility to attract not only customers but developers as well... which is what I'll call a community.

At OpenSolaris.org there is a project being done by the community called COMSTAR (Common Multiprotocol SCSI Target).  It is a clever framework which enables protocol plug-ins which speak differnent flavors of storage like Fibre Channel, iSCSI, etc.  For me it is analogous to  the old port and class driver model of my youth.  For more info on COMSTAR click here

Also see what this community member has to say. Some other notables about the storage plumbing at OpenSolaris.  The COMSTAR effort puts code into the kernel for optimization.  It will improve upon the current iSCSI Target already available and in good use today.

Open sourcing the entire storage stack implies the storage plumbing too.  For example it enables this storage stack to work for you rather than you working to pay for that expensive storage solution.

Remember Sun's OS, middleware, database and infrastructure products have the following in common-- they are models of open sourced software that more and more customers are demanding from all vendors and providers.

Thursday Dec 06, 2007

NDMP The Protocol, It's All About The Storage...

Why talk about an industry standard protocol called NDMP (Network Data Management Protocol).  I agree by itself it is a low level medium that is a check box item in the world of storage.  However it is an important piece of plumbing that enables your "house" to function.  Numerous software storage applications, particularly backup software and tape products, utilize this protocol for providing services.  While it is analogous to other storage protocols such as iSCSI and FCoE it is "the" protocol when it comes to basic backup.  Pretty sophisticated backup software choices are out there all driven by this basic protocol.  This includes multiple configuration choices such as local backup, 3-way and DAR. 

  • SUN is a working member of the SNIA community to advance the adoption of industry standards. We will be contributing source code to the SNIA effort to help update the SNIA software from NDMP v3 to NDMP v4.

  • SUN is fostering its own OpenSolaris Storage Platform community in cooperation with SNIA and other industry standards bodies to implement and enhance storage industry standards.

  • SUN plans to incorporate the SNIA software being created by the NDMP software TWG into OpenSolaris when it is approved by the SNIA membership.

The storage stack of OpenSolaris keeps getting stronger.  While NDMP is a robust service for backup, when it comes to management of massive amounts of data your mileage may vary.

 

 

 

 

 

 

 

 

While the above picture shows the NDMP protocol in application it does not speak to the entire pyramid structure on the right.  Backup can help you with the performance/cost curve of storage classes, but true storage archive management (SAM) is the one storage application that can automatically manage multiple tiers of storage according to policies crafted by the environment.   Policies are usually driven of of the metadata (the data on the data that you store) which allows you to move data between classes of storage (e.g. expensive FC disks, cheaper SATA disks and magnetic tape) based on file size, access frequency, creation date, etc.  OpenSolaris has released initial SAM/QFS code sources and will continue to do so until the entire SAM/QFS code base is available to the community.  Equally exciting are the OpenSolaris projects ADM and MMS which are bringing storage archive management to the ZFS file system.  Couple all these activities together make one extremely busy storage community working with OpenSolaris.  Given all the technologies are open source code bases you have the ability to leverage with your own code even if it is proprietary.  Fueling the opensource community benefits not only better innovation but more customer choice as well.

 

 

Friday Oct 26, 2007

FISH -n- CIFS

There is motivation when skeptics like to preach that there is no way <substitute here> is going to happen.  Examples include come from behind victories in sports, new discoveries in science and cures for disease.  Good intentions are one thing, but following through on those intentions is a bit more difficult. 

The storage community at opensolaris.org has been busy this past year with community members that include partners, industry standard bodies and numerous code contributions.  The community has been on a journey to continue to enhance, integrate, develop, share and invent technologies applicable to traditional storage and more importantly hybrid storage solutions of today and tomorrow.  The Storage Platform for OpenSolaris Distributions has just had some recent contributions.  Keep in mind these contributions to the community are free, open and not proprietary.  Code is not read only, but available for modification, improvement and bi-directional sharing according to the opensolaris licensing terms.

Recent additions of code and new opensolaris projects include:

  • Kernel based CIFS server (Build 77).  That's right-- designed in as a first class citizen of the os with a kernel based protocol, tightly coupled with NFSv4, VFS, ZFS and Active Directory.  Windows Interoperability.  Another complimentary open sourced solution along side our friends from the SAMBA community.  This service leverages the os and its capabilities.  Need infinite snapshots of your CIFS files?  Want file compression?  Strive for encryption of your data?  Not a problem.  ZFS provides these data services IN the file system.  The kernel based CIFS server source code will post here today.  Need help with source code management tools click here.

  • NDMP service.  Table stakes for backup applications.  The code will be binaries only until the SNIA working group members complete their efforts.

  • Virus Scan service.  Another service that is tightly integrated with the kernel CIFS server and the ZFS file system.  This service provides ICAP protocol support for off-board virus scan engines.

  • NFSv4 Mirror Mounts.  NFSv4 Clients can now automatically mount shared file systems on a NFSv4 Server.

The above new Opensolaris project pages will be posting in the next few weeks so stay tuned.

Adding more and more content to the storage stack of opensolaris may raise some questions of what can be done with this stack.  Well first off it enables better integrated hybrid storage solutions.  The x4500 "Thumper"with Solaris has enabled new thinking storage solutions by leveraging the hardware and software unique capabilities.  One can continue on this journey and enhance the unique hardware and software capabilities.  In fact think of it as Fully Integrated Software and Hardware on a repetitive basis.  The tighter you integrate the incremental features, the more compelling solutions with commodity components using an enterprise open sourced os one can build.  I would think that if you could provide a software heath kit of an open sourced software stack to appeal to the masses you may have something worth investigating.

Come and participate.  Opensolaris is transforming itself from open storage solutions, xVM to Solaris install revisited also known as "Indiana." 

How can you pass up FISH -n- CIFS well prepared? Help spread the word.

Thursday Aug 09, 2007

Storage as a commodity is accelerating

As company earnings continue to be announced in the 2nd half of 2007 more evidence is becoming available which points to a softening of storage demand, especially in the U.S.A.  Opinions range from the credit crunch of rising interest rates, product mix issues and full channel inventories that need to be drawn down.  While all of the above reasons may be a factor, I’d like to propose one of my own observations.

The storage industry is approaching an inflection point.  Multiple public companies are all speaking about IT consolidation and virtualization.  For newly formed companies and businesses their criteria requires designing around scalability, minimum solution costs and most importantly no vendor dependence.

I believe that companies that enjoy margins north of 50 points for storage solutions will be forced to be more competitive with storage solutions that go the way of commodity.

Let’s take as an example: block and file storage.  Solutions are provided today via a variety of proprietary embedded methods that range from traditional server/storage configurations to specialized appliances.  Storage services for all solutions typically provide support for industry standard protocols (iSCSI, NFS, CIFS, NDMP) and data services (snapshots, replication, RAID, compression, compliance, … etc.).  Storage solutions today charge you licensing fees or RTUs (Right To Use) for each protocol and data service you need.  This is a healthy revenue stream for certain storage vendors.

RAID performed by intelligent hardware controllers is a solid solution today.  However businesses are moving toward solutions based on software RAID because of enhanced protection provided by the software RAID/file system combination.  The design of the application is a factor too when multiple hardware failures can be tolerated because data copies are distributed and recoverable.  A good example of using cheap, non redundant hardware with software capable of handling multiple hardware failures is the Google File System (GFS).

Another factor at work here is a trend that can almost be defined as a default standard today—that of basing your IT solutions off of a Linux distribution.  There are multiple offerings of Linux today.  One distribution that is gaining adoption is CentOS.  Both public and private companies are deploying CentOS as an alternative to a popular Linux vendor distribution.
CentOS is built from publicly available open source SRPMS provided by a prominent North American Enterprise Linux vendor. CentOS conforms fully with the upstream vendors redistribution policies and aims to be 100% binary compatible. (CentOS mainly changes packages to remove upstream vendor branding and artwork). All perfectly legal under the terms and conditions of open source licensing.

How would the competitive landscape change if storage protocol support and data services could be added to a unix distribution? In other words take a proven enterprise unix os that is open sourced and build in additional features specific to storage.  In fact why not add other features that are becoming expected standards.  A good example here is virtulization.  If this could be done it would be an enabler for creating storage product offerings out of non proprietary software using commodity servers and storage.  This would change the current landscape especially if cost savings and prevention of vendor lock in is achievable.

There is a growing customer base that is building their IT infrastructure from hybrid storage solutions.  They don't quite fit nicely into the standard file, block or even object industry segments today.  In some ways these customers want to slightly tweak the storage, servers and software to create the hybrid.  This hybrid also turns out to be their competitive weapon.

Helping customers achieve their own individual hybrid is certainly an opportunity worth pursuing.



About

The blog of Bob Porras - Vice President, Data, Availability, Scalability & HPC for Sun Microsystems, Inc.

Search

Categories
Archives
« April 2014
SunMonTueWedThuFriSat
  
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
   
       
Today