Adventures in systems land

That got your attention didn’t it?

We’ve announced another performance and benchmark record this week, IBM WebSphere Application Server benchmark involved more than 109,850 concurrent clients and produced 14,004.42 SPECjAppServer2004 JOPS@Standard (jAppServer Operations Per Second), which translates into more than 50 million business transactions over the course of the benchmark’s hour-long runtime. That’s a lot of clients, and a lot of transactions!

The performance run was completed on IBM POWER6 BladeCenter servers powered by two dual-core IBM® POWER6® 4.0 GHz processors and IBM DB2 Universal Database v9.5 on a System p p595 running AIX.

We ran the test over 52-processors, 2-cores per processor and with SMT on. The software config included 26 WAS instances. Now, the issue here isn’t performance, 26-instances isn’t so bad from a config and deployment perspective either. But wouldn’t it be better if you could bundle that all up into a couple of racks and use cloning, automatic deployment, recovery, scheduling etc. and on an even more consolidated, energy efficient platform.

Funnily enough, we are working on that. The IBM Press release mentions IMPACT 2008, that might be good timing, I won’t be there as I’m off to do the Machu Picchu thing at the start of April.

Prior to the new WebSphere+Power double-up, the 4Q2007 record was held by Oracle on HP-UX Integrity Server Blade Cluster, with 10,519.43 JOPS over 24 server instances on 22 2-core processors; Sun also submitted a SPARC T5120 SPECjAppServer2004 benchmark with Sun Java System Application Server 9.1 running 6-nodes, 18-server instances on 48-cores, 6-chips and only scored 8,439.36 JOPS.

You can read the full press release with links to SPEC and IMPACT 2008 here.

No, not Solaris on Power, but today my long time buddy and fellow IBM Distinguished Engineer, Jim Porell, is gushing about their demo of Solaris on System z(aka the mainframe). Still no word on middleware and application vendor support. Thats when it gets interesting until then it will be a another open source and development option.

Jims’ flow can be read here. My original comment and opinion on this, here.

I initially wrote the following a version of this as a comment to John Meyers blog entry over on sun.com - Somewhere between starting the comment on his blog and finishing it, comments were closed and it didn’t get accepted.

A number of people over the past few weeks having been egging me on to respond to John’s blog entries comparing SUN and POWER offerings. It’s great being an evangelist, being the ultimate believer in a product, technology, cleaning equipment or life saving gadget, you can’t fail, the world is your oyster, your vision is world domination and your business allows you to do it, better still, they encourage you. I’m certainly not going to do a line by line analysis and deconstruction of his writings, it’s just unproductive. He has an opinion, and he is entitled to it.

Over the next few weeks though I will post some thoughts on the general assertions. Here though, is the response I originally wrote to this blog post.

John, it’s been fun reading your POWER and virtualization analysis, you are obviously passionate about your position and the technology at SUN. SUN have clearly done a good job at filling some gaps in their product portfolio over the past few years, some in response to competitive pressure from IBM and others, and some as industry leadership.

There is no doubt that SUN have done some things that have meant IBM has had to respond. However, what you seem to have glossed over, in direct comparison to POWER Systems, rather than IBM Virtualization in general, is the real need for some of the features, and their real usability, rather than just the technical implementation. Hey, but thats life through “rose-tinted spectacles”. Oh yes, this isn’t “hubris”, I didn’t create logical partitioning, but I did contribute to it as well as a number of other important virtualization technologies.

Matais, I assume you mean CTSS which was developed at MIT to run on an IBM 709 computer between 1959-1961.

One of the programmers on that project, Bob Creasey, went on to become the project lead for CP40 the first ever IBM Virtualization implementation. CTSS was really more of a time sharing system, rather than “virtualization.”

Gene Amdahl, then Chief Architect for the S/360 product line at IBM, visited MIT a number of times and had meetings with the Professor and the CTSS team with a view to making enhancements to the hardware architecture. It is reported that they didn’t see eye-to-eye over a number of things.

There is a written history and more of this than you’d ever want to know at: http://www.princeton.edu/~melinda/25paper.pdf

The concept of “domains” and logical partitions isn’t included in the above. It would not be correct though to state that Amdahl created LPARS. He actually lead a company that created a firmware/hardware implementation of multiple domains. IBM’s implementation of logical partitions differed significantly although used a similar basic premiss. Further discussion with revisions, corrections and updates probably belongs elsewhere, where it can be maintained, and not a reply to a blog post where it cannot.

There are number of companion documents that show the roll of other important users and customers which helped IBM improve its’ virtualization offerings.

Regards.

For the record I also wrote comments on the Solaris/Linux/AIX conspiracy theory here

While sitting around this evening listening to the zdnet, David Berlind and Redmonk webcast, Monkcast #12: IBM HW group OEMs Solaris to chagrin of SW group & a fly in VMware’s ointment and both on the show and a web article, a few things came up that were worth commenting on, if for no other reason than to save history from being rewritten:

Continue reading ‘Predict the future, rewrite history’

Today IBM and SUN announced that IBM would become the first Tier-1 distributor for Solaris. SUN are obviously excited about this, it gets them access to some excellent IBM System x (x86) hardware.

In the announcement, my boss, Bill Zeitler, head of IBM Systems Group also mentions that Sine Nomine and David Boyes and the effort to make OpenSolaris on the mainframe. I’ve known David for many years, back to the early Linux mainframe days and developments like this always remind me of the interesting debate we had with SUN then. David posted on the Sine Nomine web site here. I’m sure if it can be made to work effectively, David will get it done.

The first question came up over the relationship with AIX. Bill says its about custonmer choice and open interoperability in the marketplace and later when Ashley Vance from the Register asked about running Solaris on System p, the answer was we are not doing anything

Jonathon Schwartz remarked that mainframes “set the Gold Standard for virtualization with logical partitions” and “Linux support on mainframes gives customers and outstanding set of options”.Which of course is what I was saying back in 2001. The more things change, the more they stay the same.

Ashley Vance, in his write-up, I think got slightly over excited in what was little more than a comment by Bill on customer choice. The real question is what will software vendors do here? For x86 it’s not such a big deal since one Solaris x86 binary should run on any other x86 Solaris platform, but on other platforms such as System z and other non-x86 compatible platforms it’s another binary to support, more testing to do etc. Disappointed that none of the analysts thought to ask that, or for that matter if Solaris were available on POWER hardware why SUN would bother with their own UNIX hardware. Oh well

There is a replay of the IBM/SUN conference call here.

Kudos to the folks over at Sun for the new ODF plug-ins for Microsoft Office. Former colleague Simon Phipps alerted me to them in a post in his blog. Whatever I think about Suns server/hardware strategy, you can’t deny that Sun, under Simon’s direction, are doing some great things with and/for open source. Nice one!

The Plug-ins are here.

Over in Paul Murphy’s Managing L’unix blog on zdnet, he does an overly simple analysis of simultaneous Multi-threading (SMT) architecture systems versus more traditional multi-core/SMP systems and falls to at least one common misconception and misses the point technically, architecturally and operationally.

Of course I would say that…

Continue reading ‘Targeting development for future hardware’

In his del.icio.us links for yesterday, my former deptartmental colleague, Simon Phipps, who has been doing a great job as Chief Open Source Officer or what ever he wants to call his position, we like to have fun with job titles, posits that IBM should join the march to Solaris. Huh, why?

It’s funny how things worked out. When I was working with Simon in IBM SWG as their Linux technology evangelist, I was having a debate via LinuxGram with Shahin Khan, then Sun’s chief competitive officer, on our move into Linux and Suns’ obfuscation on Linux.

Sun have been doing a stellar job recently with their open source efforts and support of things like ODF, but I just don’t get why Solaris is such a big deal. For most of our systems, customer choice and the “power to leave” are delivered through a Linux implementation alongside industry standards that deliver interoperability and management of the IBM operating environment. I can only wonder how much better off the industry and customers would have been if Sun had put the same amount of time and effort into Linux in the five years that have elapsed since that debate.

Insight? Nah, just a lucky guess, who’da thunk Sun would GPL Solaris…