Montag Aug 01, 2011

Oracle Solaris Studio 12.3 Beta Program

The beta program for Oracle Solaris Studio 12.3 is now open for participation.  Anyone willing to test the newest compiler and developer tools is welcome to join.  You may expect performance improvements over earlier versions of Studio as well as GCC that make testing worth your while.

Happy testing!

Donnerstag Jul 28, 2011

No humor left?

Just in case we all forgot that originally, Unix people were blessed with lots of humor:
OTN ID 1172413.1: How to make a grilled cheese sandwich

Dienstag Jul 05, 2011

Some Thoughts about the new SPARC Roadmap

As most of you know by now, Oracle has recently released an updated roadmap for its SPARC based servers.  Of course, others also publish roadmaps, and we all know how CPUs tend to slip and roadmaps sort of evolve backwards over time.  So what's the value in this new roadmap?

When Oracle aquired Sun, there were all these big promises about increased investement in SPARC and Solaris.  We've already seen the first delivery on these promises:

  • The SPARC T3 CPU doubled the throughput per socket for the T-Series line of servers.
  • Solaris 11 Express gives customers a supported (!) preview of what's coming with Solaris 11
  • M-Series servers can be upgraded yet again for up to an additional 20% performance gain.

All this is shown in the new roadmap as tick-off items.  All this was already well into completion at the time of the aquisition, although the speed with which T3 came to market can already be attributed to Oracle's accelerated investment.  Nevertheless, the real prove point will be the next promise on that roadmap.  It says: "3x Single Strand" for a T-Series CPU.  By now, we all know that the name will probably be "T4".  Rick Hetherington told us a while ago, that the chip will have 8 cores and will execute single threaded workloads up to 5 times faster than T3.  That would be even more than the 3x promised by the roadmap.  The only question that remains is:  Will Oracle deliver?

While no one can tell until the chip and systems are actually announced, chances are good it will.  Why else would Oracle invite customers to participate in a T4 Beta Program?

Back to the original question:  What's the value of this new roadmap?  Well, the foundation of trust, based on delivering on promises, looks good for that part of the roadmap arrow that's already past.  Oracle delivered, and it delivered on time.  That's more than some others can claim.

A little disclaimer to keep everyone happy:  This is my personal view and not an official statement from Oracle.

Donnerstag Jun 09, 2011

OVM Server for SPARC 2.1 is here!

The newest version of OVM Server for SPARC aka LDoms is released!  Here's the press release...


What, already a new version again?  Well, the most missed feature in the previous versions was finally completed, and we didn't want to keep everyone waiting ;-)  The new version 2.1 turns "Warm Migration" into "Live Migration".  All the other improvements can be found in "What's New".  Once I have further details about Live Migration, I'll post them here.  You can find the download on MOS and the documentation on OTN.

Mittwoch Jun 08, 2011

Erasing disks securely

Actually, both the question and the answer are old and well known.  However, these things tend to be forgotten and pop up as questions from time to time.  Hence a little reminder for all of us:

Solaris makes it easy to erase a disk so that all the data can't be restored, even with sophisticated methods.  There is a subcommand "analyze/purge" in the command format(1M) that does it all for you.  It will overwrite the selected area of your disk (usually s2) a total of four times with different patterns to achieve this.  Of course, depending on the size of the disk, this might take a while.  But it's secure enough to comply with Department Of Defence(DOD) wipe disk standard 5220.22-M.

Some more details are here:

Note that this method does not apply to SSDs of all kind!  And of course, to avoid any risk of losing your data with your disk, simply encrypt it!  It's quite easy using ZFS or Oracle TDE :-)

Dienstag Mai 24, 2011

ILOM for ALOM Users

ALOM ist dead, long live ILOM! The current T3 systems use the new ILOM 3.0.  This version no longer supports the legacy ALOM commands.  That's a change for all those that got used to the alom command syntax over the years.  A table comparing the old and new commands would be helpful.  But where would one find such a table?  In the dokumentation!

But it's not always easy to find what you're looking for right away. To help, here's a link to exactly that table:

ALOM ILOM Commando Comparison

Dienstag Apr 26, 2011

Moving

This blog will move.  Starting May 6th 2011, the new address will be http://blogs.oracle.com/cmt/  Everything else will remain unchanged :-)

Donnerstag Mrz 31, 2011

What's up with Solaris 11?

Interested in the upcoming Solaris 11?  What will be the highlights?  What exactly is the new packaging format, how does the new installer work?  What do the analysts think?


All this will be covered in the Solaris Online Forum on April 14, starting at 9 am PST.  This will be a live event where you can ask questions. (A recording will be available afterwards.)  Speakers are all high level members of development and product management.


All further details can be found at the registration page.

Freitag Jan 28, 2011

Cash for Clunkers 2.0

This ad is too nice to not repeat it here... Enjoy!
Of course this is quite serious: Trading in an HP Superdome for an M8000 or M9000 is indeed a deal you can only win!

Mittwoch Jan 26, 2011

Logical Domains - sure secure

LDoms Oracle VM Server for SPARC are being used wide and far.  And I've been asked several times, how secure they actually were.  One customer especially wanted to be very very sure. So we asked for independent expertise on the subject matter.  The results were quite pleasing, but not exactly night time literature. So I decided to add some generic deployment recommendations to the core results and came up with a whitepaper. Publishing was delayed a bit due to the change of ownership which resulted in a significant change in process.  The good thing about that is that now it's also up to date with the latest release of the software. I am now happy and proud to present::


Secure Deployment of Oracle VM for SPARC


A big Thanks You to Steffen Gundel of Cirosec, who laid the foundation for this paper with his study.


I do hope that it will be usefull to some of you!


Enjoy!

Freitag Dez 17, 2010

Solaris knows Hardware - pgstat explains it

When Sun's engineering teams observed large differences in memory latency on the E25K, they introduced the concept of locality groups (lgrp) into Solaris 9 9/02. They describe the hierarchy of system components, which can be very different in different hardware systems. When creating processes and scheduling them onto CPUs for execution, Solaris will try to minimize the distance between CPU and memory for optimal latency. This feature, known as Memory Placement Optimization (MPO) can, depending on hardware and appliation, significantly enhance performance.

There are, among many other things, thousands of counters in the Solaris kernel. They can be queried using kstat, cpustat, or more widely used tools like mpstat or iostat. Especially the counters made available with cpustat depend heavily on the underlying hardware. The it hasn't always been easy to analyze the performance benefit of MPO and the utilization of individual parts of the hardware using these counters. For cpustat, there was only a perl-script called corestat to help understand T1/T2 core utilization. This has finally changed with Solaris 11 Express


There are now three new commands: lgrpinfo, pginfo und pgstat.

lgrpinfo shows the hierarchy of the lgroups - the NUMA-architecture of the hardware. This can be useful when configuring resource groups (for containers or standalone) to select the right CPUs.

pginfo shows a different view of this information: A tree of the hardware hierarchy. The leaves of this tree are the individual integer and floatingpoint unit of each core.  Here's a little example from a T2 LDom configured with 16 strands from different cores:


# pginfo -v
0 (System [system]) CPUs: 0-15
|-- 3 (Data_Pipe_to_memory [chip]) CPUs: 0-7
| |-- 2 (Floating_Point_Unit [core]) CPUs: 0-3
| | `-- 1 (Integer_Pipeline [core]) CPUs: 0-3
| `-- 5 (Floating_Point_Unit [core]) CPUs: 4-7
| `-- 4 (Integer_Pipeline [core]) CPUs: 4-7
`-- 8 (Data_Pipe_to_memory [core,chip]) CPUs: 8-15
`-- 7 (Floating_Point_Unit [core,chip]) CPUs: 8-15
|-- 6 (Integer_Pipeline) CPUs: 8-11
`-- 9 (Integer_Pipeline) CPUs: 12-15

As you can see, the mapping of strands to pipelines and cores is easily visible.

pgstat finally, is a worthy successor of corestat. It gives you a good overview of the utilization of all components. Again, an example, on the same LDom, which at the same time shows almost 100% core utilization, something I don't find very often...


# pgstat -Apv 1 2
PG RELATIONSHIP HW UTIL CAP SW USR SYS IDLE CPUS
0 System [system] - - - 100.0% 99.6% 0.4% 0.0% 0-15
3 Data_Pipe_to_memory [chip] - - - 100.0% 99.1% 0.9% 0.0% 0-7
2 Floating_Point_Unit [core] 0.0% 179K 1.3B 100.0% 99.1% 0.9% 0.0% 0-3
1 Integer_Pipeline [core] 80.0% 1.3B 1.7B 100.0% 99.1% 0.9% 0.0% 0-3
5 Floating_Point_Unit [core] 0.0% 50K 1.3B 100.0% 99.1% 0.9% 0.0% 4-7
4 Integer_Pipeline [core] 80.2% 1.3B 1.7B 100.0% 99.1% 0.9% 0.0% 4-7
8 Data_Pipe_to_memory [core,chip] - - - 100.0% 100.0% 0.0% 0.0% 8-15
7 Floating_Point_Unit [core,chip] 0.0% 80K 1.3B 100.0% 100.0% 0.0% 0.0% 8-15
6 Integer_Pipeline 76.4% 1.3B 1.7B 100.0% 100.0% 0.0% 0.0% 8-11
9 Integer_Pipeline 76.4% 1.3B 1.7B 100.0% 100.0% 0.0% 0.0% 12-15
PG RELATIONSHIP HW UTIL CAP SW USR SYS IDLE CPUS
0 System [system] - - - 100.0% 99.7% 0.3% 0.0% 0-15
3 Data_Pipe_to_memory [chip] - - - 100.0% 99.5% 0.5% 0.0% 0-7
2 Floating_Point_Unit [core] 0.0% 76K 1.2B 100.0% 99.5% 0.5% 0.0% 0-3
1 Integer_Pipeline [core] 79.7% 1.2B 1.5B 100.0% 99.5% 0.5% 0.0% 0-3
5 Floating_Point_Unit [core] 0.0% 42K 1.2B 100.0% 99.5% 0.5% 0.0% 4-7
4 Integer_Pipeline [core] 79.8% 1.2B 1.5B 100.0% 99.5% 0.5% 0.0% 4-7
8 Data_Pipe_to_memory [core,chip] - - - 100.0% 99.9% 0.1% 0.0% 8-15
7 Floating_Point_Unit [core,chip] 0.0% 80K 1.2B 100.0% 99.9% 0.1% 0.0% 8-15
6 Integer_Pipeline 76.3% 1.2B 1.5B 100.0% 100.0% 0.0% 0.0% 8-11
9 Integer_Pipeline 76.4% 1.2B 1.5B 100.0% 99.8% 0.2% 0.0% 12-15

SUMMARY: UTILIZATION OVER 2 SECONDS

------HARDWARE------ ------SOFTWARE------
PG RELATIONSHIP UTIL CAP MIN AVG MAX MIN AVG MAX CPUS
0 System [system] - - - - - 100.0% 100.0% 100.0% 0-15
3 Data_Pipe_to_memory [chip] - - - - - 100.0% 100.0% 100.0% 0-7
2 Floating_Point_Unit [core] 76K 1.2B 0.0% 0.0% 0.0% 100.0% 100.0% 100.0% 0-3
1 Integer_Pipeline [core] 1.2B 1.5B 79.7% 79.7% 80.0% 100.0% 100.0% 100.0% 0-3
5 Floating_Point_Unit [core] 42K 1.2B 0.0% 0.0% 0.0% 100.0% 100.0% 100.0% 4-7
4 Integer_Pipeline [core] 1.2B 1.5B 79.8% 79.8% 80.2% 100.0% 100.0% 100.0% 4-7
8 Data_Pipe_to_memory [core,chip] - - - - - 100.0% 100.0% 100.0% 8-15
7 Floating_Point_Unit [core,chip] 80K 1.2B 0.0% 0.0% 0.0% 100.0% 100.0% 100.0% 8-15
6 Integer_Pipeline 1.2B 1.5B 76.3% 76.3% 76.4% 100.0% 100.0% 100.0% 8-11
9 Integer_Pipeline 1.2B 1.5B 76.4% 76.4% 76.4% 100.0% 100.0% 100.0% 12-15

The exact meaning of these values is nicely described in the manpage for pgstat, so I'll leave the interpretation to the reader. With this little tool, performance analysis, especially on T2/T3 systems, will be even more fun ;-)

Mittwoch Nov 24, 2010

Encrypting Your Filesystem with ZFS and AES128

ZFS filesystem encryption is finally available in Solaris 11 Express.  This closes a gap in Solaris that hurt all those that carried their data around with them.  But of course there are many good reasons to encrypt data living on disks well secured in a datacenter.  After all, they will all leave the datacenter in one way or another eventually...

Enough introduction, here's how simple this is:

  1. You will need to upgrade the zpool intended to host the encrypted filesystem to version 30.  Issue a simple "zpool upgrade <poolname>.  Of course, you can skip this step on a newly installed Solaris 11 Express.
  2. Now create a new filesystem, with encryption enabled: zfs create -o encryption=on <poolname/newfs>
    The command will interactively prompt for a passphrase which will be used to generate the key for this filesystem.  You're done!  You can not encrypt an already existing filesystem.  Of course there are several more options on how and where to store the key.  Just have a look at the manpage :-)

Likewise, you also have a choice of three different key lengths for AES, the algorithm used for encryption.  The default used for "encryption=on" is AES-128 in CCM mode.  But you can also choose the longer 192 or 256 bit keys.  While developing ZFS crypto, it was discussed what default keylength to choose.  AES-128 was chosen for two reasons:  First, of course, the 128 bit variant is faster than the longer key lengths, especially without hardware acceleration like it is available in the SPARC T2/T3 and Intel 5600 Chips.  Second, there is new research including successful attacks on AES256 and AES 192 that requires a search of only 2\^39.  These attacks don't work for AES128, which is therefore, as of today, not only faster, but also more secure than the variants with longer keys.

More details about ZFS Crypto in the ZFS Admin Guide.

Freitag Nov 05, 2010

DOAG 2010

This years DOAG 2010 (German Oracle User Group) Conference will see me busy.  I'll not only present about cryptography and new Oracle hardware, but also answer any questions you might have at the exhibition.  Hope to meet you there!

Donnerstag Okt 28, 2010

T3 twice as fast or half as expensive?

I don't usually copy other's work.  But this time, "The Register" was kind enough to do all the tedious price and performance comparisons.  So, thankful for not having to do this myself, I point to their latest article.  It discusses and shows T3 price and performance compared to the earlier T2 systems.  All that's left to ask is: Is T3 twice as fast or half as expensive?  What a nice choice!

Dienstag Sep 21, 2010

No Excuse for no Security

Ever since Sun shipped the UltraSPARC T2 CPU, there was no excuse for not using SSL security for web services.  With the new T3 chips, this is more true than ever.  Not only have the supported algorithms been modernized.  The required documentation on how to use this feature has also been updated.  Find the first two papers hers.  I expect there to be more soon.


Happy encrypting!

About

Neuigkeiten, Tipps und Wissenswertes rund um SPARC, CMT, Performance und ihre Analyse sowie Erfahrungen mit Solaris auf dem Server und dem Laptop.

This is a bilingual blog (most of the time). Please select your prefered language:
.
The views expressed on this blog are my own and do not necessarily reflect the views of Oracle.

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