The IT Business

I thought a while just write another comment about the HP785G6 benchmark in the comment section of the "Case Design" or the Do more with less" article. At the end i thought, this will be a fine article.

At first: Yes, the benchmark of Sun used 256 GByte, whereas as the HP 128 GByte. At first this sounds like benchmark not comparable to each other. One of the basics in SoHo benchmarks is to compare benchmarks with the same memory size. The problem: You can't use this knowledge here.

But: The amount of memory just matters in SAP-SD when you don't have enough. When you don't believe me, ask your favorite vendor of IT equipment and ask for SAP-SD numbers with different memory sizes. The answer will be similar. To be more exact, the speed of the memory has a bigger impact than the size of the memory, given that you have enough memory.

Another reader summarized it in a good manner: If HP could get a better result just by adding memory, they would have done it. It's difficult to increase compute power when you are already using the fastest CPU. It's difficult to increase the core count when you are already using the the CPU with the highest number of cores. But just plugging more memory into the system is easy. I haven't heard any news about HP having a shortage of memory modules.

But it think there is a different reason, why HP just used 128 GB. When you look at page 15 of the quick specs pdf of the HP DL785G6 speed you will find the following note:

When only PC2-6400 DIMMs modules are installed with a processor then memory bus speeds for 4 or fewer, 6 or 8 DIMMs per processor will operate at PC2-6400, PC2-5300 and PC2-4200 respectively. All other processor and memory configurations will operate at PC2-5300 with 4 or fewer DIMMs and PC2-4200 with

The largest memory flavor with PC2-6400 is "8 GB REG PC2-6400 2 x 4 GB". The DL785 has 64 DIMM slots. To keep the memory bus at PC6400, you can just populate 32 of them. 32x4=128GB. They've ran the system at the top configuration that allowed them to use faster DIMMs. But you just to that, when you have enough memory in the smaller configuration, because the impact of having not enough memory is much larger than the impact of 133 MHz more memory speed.

From this point it get's even more strangely: The X4640 didn's just had slower a slower CPU (200 MHz), it has also a tad slower memory (a X4640 use PC2-5300 for all memory configurations). This makes the difference between the systems even larger as suggested by the pure SAPS numbers.

For the database something similar is valid. Roundabout 3-4% percent of an SAP-SD benchmark are database-bound and basic math suggests, that even a database 25% faster has a rather small impact on the database.

Of course it would be nicer to have an exact proof by having benchmark results just differing by a single point but we have to live with it, that such results aren't available.

But Sun is in a nice situation here, when you look at the situation with some logic. The vast difference between the HP DL785 and the X4640 can be explained by three major points:

• The X4640 is vastly better than the DL785.
• Solaris is vastly better than Windows/Linux.
• The benchmarking team at HP is just outright incompetent.

I think we can say that the third point can be erased from the list. I know HP as a competent company. As usual the an organisation of such a size is a victim of the bell curve, thus it has it share of dumb asses and outright geniuses. So your personal experience may differ, but let's just assume that HP is competent enough to tune a SAP.

So we are at:

• The X4640 is vastly better than the DL785.
• Solaris is vastly better than Windows/Linux.
• The benchmarking team at HP is just outright incompetent.

Leaves us with "better OS" or "better HW". Either way it's nice for Sun and that's exactly the reason why we won't see an Solaris/DL785/Oracle result from HP. Such a benchmark could have two outcomes: The DL785 yields the same result than the X4640 or it delivers worse result. The first would show that Solaris is a better operating system than Linux, the second outcome would show that the X4640 is a better system. Both is positive for Sun, both outcomes are negative for HP.

When the system is 25% slower, this would really kill the DL785. If Solaris is 25% better, then this would kill all the arguments that Linux is as scaleable as an enterprise Unix. If it's a combination of both, well ... then this just proofs that Sun has a more competitive package. I'm fine with all three outcomes.

Of course you could ask "Why does Sun don't deliver an Windows/Linux result?". Well ... honestly: Would you believe us? We have our own operating system and i'm sure the prefered excuse when the benchmarks with Linux or Windows are made by Sun would be "Sun didn't tune Linux correctly, they just want to make Solaris looks good". Of course Sun won't make up the results, as this is too easy to detect and to confirm and would habe an horrible impact PR-wise. And at the end it's HPs task to show that their system is comparable, not ours.

While preparing a presentation, i just found a good example, why i prefer the Sun case design over the one of HP.

This is an HP DL785G6 from the top (this one is from an HP video):


This is the Sun X4640 from the top:

There is an interesting article about the Moglen Letter at Groklaw inclusive a really disgusting reaction of Mr. Florian Müller (as we tend to say in german: I can't eat enough to vomit the amount i would like to) . But read it yourself.

Reuters reports in "REFILE-Activist lawyer sees flaw in EU Oracle-Sun report" about an analysis of the EU report in regard of the Sun/Oracle deal made by Eben Moglen

Moglen said that the EU document contains factual errors in its analysis of the role MySQL's licensing terms plays in securing competition in the software industry.

and

"The issues raised (by the commission) concerning the GPLv2 status of the MySQL code base do not warrant a conclusion that this transaction threatens significant anti-competitive consequences," Moglen told EU regulators in a Nov. 19 letter.

Just for the readers unaware of Mr. Moglen - he founded the SLFC:

The Software Freedom Law Center (SFLC) is an organization that provides pro bono legal representation and related services to not-for-profit developers of free software/open source software. It was launched in February 2005 with Eben Moglen as Chairman. Initial funding of $4 million USD was pledged by Open Source Development Labs.

Andreas Stiller wrote in his "Prozessorgeflüster" column (the last remaining island of constant high quality in the c't) about a nasty in the Intel Nehalem. For i7 900 you will find the errata at "AAJ121. Unexpected Interrupts May Occur on C6 Exit If Using APIC Timer to Generate Interrupts" in the pdf Intel® CoreTM i7-900 Desktop Processor Extreme Edition Series and Intel® CoreTM i7-900 Desktop Processor Series. For other Nehalem procs you will find matching reports under AAK119 for Xeon 5500, AAM123 for Xeon 3500, AAO89 for Xeon 3400 and AAN87 (Core i7-800, Core i5). That's not that interesting ... bugs in CPUs are quite common

However Intel thought, that this bug isn't triggered by any commercially available software.Well ... not exactly. There is one: Windows Server 2008 R2 with enabled Hyper-V. An hot fix is available, but the interims workaround is really interesting: Switching those C-states and thus power-saving. But you will kill the Turbo Boost feature as well with such a step, as Turbo Boost is the other side of the medal of the Power Saving feature.

But that isn't still the most interesting point of this bug. The most interesting part is a atmospheric one. The relation between Intel and Microsoft must have experienced a cool down recently. Andreas Stiller writes in his article:

And it gets worse: the editorial office received a non-published raw version of the error report that even included the brusque “preferred solution” of not using the processors in question. Strong stuff: Microsoft advises against the application of Nehalem processors in servers – so Intel took to the barricades and was able to prevent Microsoft publishing this advice at the very last second.

I would have really liked to see the reaction of the usual news outlets and the market, when Intel would have not been able to convince Microsoft to change the report

While waiting for my tea water boiling, i thought that it's great to work in IT. Hypes are recombinating faster than the newest flu strain. So you can hype the same stuff again and again: Combine grid computing with a credit card system and call it Cloud Computing. And you have stuff to about talk for hours again

The European Commission has expressed their objections about the Sun acquisition by Oracle. Sun notified the SEC with an 8K filing available at EDGAR

It's based on the opinion, that the purchase of Mysql would harm the database market. I've already expressed my strong feeling, that this is utter nonsense. I think it's unnecessary to repeat it. I don't have any understanding for this decision, as it doesn't help competition and customers. In my opinion, it just helps some interested parties.

Oracle answered with a press release to the filing of the objections:

The Commission's Statement of Objections reveals a profound misunderstanding of both database competition and open source dynamics. It is well understood by those knowledgeable about open source software that because MySQL is open source, it cannot be controlled by anyone. That is the whole point of open source.

Even the Department of Justice issued a press release regarding this EC decision:

"Several factors led the Division to conclude that the proposed transaction is unlikely to be anticompetitive. There are many open-source and proprietary database competitors. The Division concluded, based on the specific facts at issue in the transaction, that consumer harm is unlikely because customers would continue to have choices from a variety of well established and widely accepted database products. The Department also concluded that there is a large community of developers and users of Sun's open source database with significant expertise in maintaining and improving the software, and who could support a derivative version of it.

However the cards are on the table now and the EC should tell Oracle and Sun, what they want to solve those objections. So it's a good that they filed the objections now and not at the last possible moment.

The New York times reports about an lawsuit filing against Intel in "2nd UPDATE: NY AG: Dell Got $6B Via Secret Intel Pact >DELL" :

Dell, the world's third-biggest computer maker based on shipments, was allegedly paid about $6 billion between February 2002 and January 2007, according to the lawsuit. In one fiscal quarter, the lawsuit says payments from Intel constituted 116% of Dell's reported net income.

Interesting news as this sheds a different light to the financial performance of Dell and you could just speculate about the margins of Intel that allow them to pay $6B and still making money.

You will find the complaint for further reading at the site of the Attorney General of the State New York. The pdf of the complaint is available here.

On 16th December i will talk about Hadoop on Sun at the Apache Hadoop Get Together in Berlin. The talk will be divided in three parts:

• Advantages of using Hadoop on CMT systems like the T5120
• Using Hadoop in conjunction with Solaris Features like Container and Resource Management in non-dedicated architectures
• Hadoop and Grid Engine

There will be short live demonstrations in the second and third part of the presentation.

My presentation will not be the only one. Richard Hutton from nugg.ad talks about data processing with hadoop in his presentation "Moving from five days to one hour." Nikolaus Pohle from Nurago talks about data analysis with Hadoop for online market research.

I wrote at several occasions, that this "Oracle purchase of Sun will limit database competition" isn't true as you can easily fork mysql and do your own development. Matt Asay has the same opinion and points to recently launched service by Amazon to substantiate his opinion:

The European Commission must be feeling a bit silly right about now. Despite insisting that Oracle has not responded to its requests for comment and concessions in its planned acquisition of Sun Microsystems (and the open-source database MySQL), Amazon.com recently offered the EC all the proof it needs that MySQL competition remains alive and well.[....]Oracle hasn't even started with MySQL yet, and it already faces significant competition, not to mention the other MySQL forks (e.g., Drizzle).

I think the steps of Amazon are a nice proof for the successful forkability of Mysql. Take the stuff, make a business out of it and make money. The more i think about this while situation, the more i think it's just a interest of a single person that blocks a merger that is good for everyone in IT. Nobody can be interested in IBM getting stronger (okay, of course: beside IBM), due to the FUD surrounding the situation of the Sun/Oracle merger. But this is the consequence at the moment.

There is a new overview/FAQ document regarding the planned Oracle/Sun merger available at oracle.com. I will cite some key passages in this article.

In my humble opinion the objective of this "Oracle shouldn't get Mysql" is pretty clear now. The Register reports about an discussion with Florian Müller in "Sun's MySQL fork survival theory ripped".

It isn't about the source. It isn't about the technology. It isn't about competition. It's about the brand.

"The value of the brand is huge - that's why the valuation is so great. MariaDB [the MySQL fork created by founder Monty Widenius] - it would take a long while to get there."

I assume, that Mr. Widenius has recognized that that MariaDB is an horrible name for marketing and wants the "Mysql" back. The critics of the Oracle/Sun merger don't want to establish a new name. Some of the proponents just saw that you can't sell the cake and eat it, too.

I think the article "MYSQL FOUNDER OUTLINES SOLUTION: INSTEAD OF LETTING SUN SUFFER, ORACLE SHOULD SELL MYSQL (ORCL, JAVA)" from Mr. Widenius is really interesting:

In order to support the regulators' work on the case, Widenius' new company, Monty Program Ab, works closely with Florian Mueller, a MySQL and EU affairs expert. Widenius said: "Florian gave MySQL strategic advice from 2001 on and was a shareholder until the sale to Sun in 2008, and with our support led an award-winning campaign against a proposed EU law on software patents. In August he helped us to demonstrate to the EC the need to investigate this merger and he is now on board again to meet the information needs of regulators, journalists and analysts."

In my opinion, Mr. Mueller is not much more than a lobbyist for the interests of Monty Programm Ab. At least this press release reads this way.

I can imagine how the proposed solution should look like: Oracle wants to stop loss, sell Mysql for a few million back to Monty AB (he has enough money due to the acquisition of Mysql by Sun) and Mr. Widenius has perfected the eat and sell cake.

Sorry ... Sun paid a large heap of money to purchase Mysql ... not only for this code, but for the brand, too. As i assumed before, it's just the money that surrounds the mysql ... it's not about Mysql and it's development. "Hey ... selling it to you was a big mistake, we want it back ... but hey ... we keep the money". It doesn't work that way. This behaviour isn't that far away from blackmailing: "I was able to block something important for you, and i will only step out of the way, when i get what i want". Disgusting. They play with 30.000 employees. But i'm pretty opinionated, that Larry shouldn't give in to such a behavior. Really, really disgusting.

By the way: Establishing a new fork is really simple because of the way, most Mysql installations were distributed to the customers. Just get Fedora, Debian, Suse and let's say Ubuntu into the boat when you fork Mysql and the branding is no issue at all. Because most of the Linux admins doesn't search for their packages at the internet ... they just ask their package manager.

The news about the removal of ZFS from Sourceforge resulted in a reinvigorated interest in the NetApp/Sun lawsuit. A multitude of web pages reported about this step. There is a comment from an authoritative source about the "Why?". I will not add something to that.

But i want to say something to the ever ocurring references to this lawsuit: Well ... i have to admit, that this lawsuit wasn't in my focus over the last few months, as i thought "Let the lawyers do their job, this can take a while". But with all this comments a short article looking into the developments since the beginning of this year looked as a necessary choice.

There is one point that i find really interesting: Everyone is talking about the point that Sun may loose ZFS, but out of a strange reason most people aren't talking about the point that NetApp has a similar problem by the countersuing, as they can lose core patents behind WAFL. And they have already loosed some patents or at least important claims.

A source of information is the ZFS Sun/NetApp lawsuit page at Sun. Let's just take the published PTO documents and not the comments by Sun.

To dodge the countersuing, NetApp tried to invalidate some patent (5761662 ("Personalized information retrieval using user-defined profil") for example) but had no success. The PTO send an "Notice of Intent to Issue" after the reexamination. The same is valid for 6591303 ("Method and apparatus for parallel trunking of interfaces to increase transfer bandwith").

On the other side:

• Many claims of NetApps patent 5819292 ("Method for maintaining consistent states of a file system and for creating user-accessible read-only copies of a file system") has been rejected by the PTO in a Final Office Action.
• The PTO issued an non-final office action for Patent 7174352 ("File system image transfer") deleting all claims that have been disputed by Sun.
• In December 2008 all claims of patent 7313720 ("Technique for increasing the number of persistent consistency point images") has been rejected by the PTO as stated in this non-final office action

At the moment i don't think that Sun will get problems in the NetApp/Sun lawsuit. I see more problems for NetApp when core patents are invalidated. However: At court and at sea you are in god's hand.

Jeff Bonwick writes about this situation:

Apple can currently just take the ZFS CDDL code and incorporate it
> (like they did with DTrace), but it may be that they wanted a "private
> license" from Sun (with appropriate technical support and
> indemnification), and the two entities couldn't come to mutually
> agreeable terms.
I cannot disclose details, but that is the essence of it.

Jeff

The discussion surrounding Mysql is somewhat depressing. At the moment it's a major road block on the way to the merger of Sun and Oracle. When you read all this stuff in the public discussion, you could come to the conclusion that Sun is just consisting out of Mysql. In my opinion it doesn't deserve this role.