HP Announces First Linux Cluster TPC-C Benchmark with Oracle and Red Hat

HP today announced that HP ProLiant servers have delivered the first Linux TPC-C benchmark results running Oracle 9i Real Application Clusters on the Red Hat Linux Advanced Server operating system. With this benchmark, HP ProLiant servers become the first industry-standard server platform to offer enterprise-class performance for a clustered Oracle database in a Linux environment.

Demonstrating the cost and maintenance benefits of running Linux-based hardware and software in enterprise-operating environments, an 8-node cluster of HP ProLiant DL580 servers using Intel® Pentium® III Xeon processors and HP StorageWorks MSA1000 storage system achieved 138,362.03 tpmC (transactions per minute) at a cost of $17.21/tpmC with Red Hat Linux Advanced Server.

TPC is a non-profit corporation founded to define transaction processing and database benchmarks and to disseminate objective, verifiable TPC performance data to the industry.

The results demonstrate that HP ProLiant servers currently offer the best price/performance ratio for companies choosing to deploy Oracle9i Real Application Clusters on Linux. Complete benchmark results and details are available at .

"HP is committed to working with leading partners, including Oracle and Red Hat, to bring proven, enterprise-ready deployment solutions for Linux to market," said Rick Becker, vice president, Operating System Alliances, and chief technology officer of software, HP Industry Standard Servers.

"Benchmark testing of this caliber requires commitment, experience and considerable resources -- all of which HP is dedicated to providing to validate Linux-based solutions for our industry-standard ProLiant servers.

The results of the benchmark prove that HP delivers one of the most cost-effective, high-performance, and manageable Linux-based database solutions available today."

"With Linux demonstrating that it's an attractive enterprise computing option, companies are looking to deploy proven applications for Linux," said Doug Kennedy, vice president, Platform Alliance, Platform Technologies Division, Oracle Corporation. "Through our strong engineering partnership with HP and extensive testing of our combined solutions, Oracle can now deliver our flagship database solutions on industry-standard HP ProLiant servers running Red Hat Linux Advanced Server with the same levels of reliability, scalability and performance enterprises have come to rely on from other operating systems."

"The benchmark on HP ProLiant servers demonstrates that Red Hat Linux Advanced Server is ready for robust, high-performance, highly available, enterprise-level applications," said Paul Cormier, executive vice president of engineering, Red Hat. "Red Hat will continue to work with leading partners, including HP and Oracle, to test and demonstrate the viability of running enterprise-ready applications on Red Hat Linux Advanced Server."

HP parallel database cluster solutions for Oracle 9i Real Application Clusters on Red Hat will be available soon as fully pre-installed solutions (hardware, software and installation included) via HP integration partners or as a custom-designed solution direct from HP.

HP has more servers certified for Red Hat Linux Advanced Server than any other technology provider. Information about additional HP ProLiant servers certified for Red Hat Linux Advanced Server is available at http://hardware.redhat.com/hcl/.

Wow! MS pulls a better score on faster hardware! Wow! But running on identical hardware, Linux runs faster, as shown by the chart. Gee, I wonder what will happen when they put Linux on the same hardware as the number one scoring system? Nice spin attempt, but once again, your FUD doesn't stand up to the light of day. Back into your crack, cockroach, the lightswitch is on.

Well, considering that the Linux box that you're crowing about won't even be available for another six months, I expect that we won't know for the next year whether Linux can match what Win2K does today. Nice try.

Well, considering that the Linux box that you're crowing about won't even be available for another six months, I expect that we won't know for the next year whether Linux can match what Win2K does today. Nice try.

And how long did it take for the Windows machines to go from development to shipping?

I notice that once I debunked your "it's only at #5" FUD you have to switch gears and find some other nit to pick.

But many people act like that when they feel threatened.

And Linux is indeed a threat to Microsoft. It won't happen today, or even in six months, but it will happen. Microsoft is going to be dethroned.

The leading indicators are that it will be Linux that dethrones them. The evidence keeps piling up. You can keep burying your head in the sand and spewing your FUD or you can admit it and pile onto your heros in Redmond to do something about it.

The business model that Microsoft has so heavily invested in, namely sell garbage, take customer complaints under advisement and sell more garbage, is about to implode. Microsoft just can't compete with free, high-quality software.

You better hope so. After all, you're betting your career on it. And if you're wrong...

And if I'm wrong, I'll have to move to another platform. Since I bounce between BSD, Linux, Solaris, AIX and MacOS already, a lateral move to a new, up-and-coming operating system won't be such a big deal for me. And I regularly check out new systems as they become available, so I'll probably have a passing aquaintance with whatever OS displaces Windows.

You, on the other hand, being tied to a legacy OS, are going to be in real trouble. I suggest you start studying the habits of the Amiga users. Windows experts will be about as relevant as they are.

More than likely, I'm going to retire on my sailboat and laugh at you, either way.

Only if you get your retirement money out of Microsoft-dependent stocks.

Let's see, if you see it coming and pull your money out but keep pretending that you are Microsoft's ad agency around here, you're a hypocrit.

If you leave your money it, you'll be broke.

I see tough times ahead for you, my friend.

Wow! MS pulls a better score on faster hardware! Wow!

The MS solution is tested on a ProLiant DL760-900-128P with 272 processors.

Total system cost: $10,603,803.

The Linux solution uses the HP ProLiant DL580-PDC 32P C/S , 32 processors.

Total system cost: $2,380,546.

So let's see . . . so if you're willing to spend a mere $8 MILLION more per machine, you can get better performance out of Windows.

Based on your numbers, you can spend roughly 39k per processor to get 700k transactions per minute, or you can spend 74k per processor to get 140k transactions per minute. Which option sounds better, especially if you need at least 500k tpm, today?

The better option, to me, sounds like to use Linux, which you can set up today on a honkin machine if you so choose. At every level tested, the Linux solution is cheaper and more stable than the MS solution. There's no reason to assume this will change, especially since Linux has proven to scale better than Windows at lower levels.

Would you rather be married to Microsoft or Oracle?

That's easy, the Oracle/Linux approach. With SQLServer/IIS/Windows you're locked into a single vendor for all products, and that vendor is known for stability problems. And expect to be forced into an upgrade path -- .NET -- that will lock you in even tighter to that single vendor. The one SQLServer db I have to deal with creates 80% of our DB problems.

With Oracle/Linux, you have a wide variety of choices of Linux vendors, a wide variety of web server choices, and Oracle is perhaps the most stable big DB around. The solution is more scalable without a doubt.

The single biggest thing to avoid as a businessperson is 'single vendor lockin'. Because then that vendor owns YOU. You're at the mercy of a single corporate entity. For example, if they decide the next version of their product -- like Longhorn -- isn't going to be backward compatible and you have to pay for all new copies of the same software, you don't have much choice.

Consider:

If you just went with 4 of the Linux clusters tested here, total cost $9,522,184, would yield you 553,348 tpc.

You're well over 500 tpc, and saved over a million dollars.

And have a more stable OS, and a far more stable DB.

And that's not even going with a more honkin machine. Going up on the hardware is likely to provide even more cost savings and better performance gains.

I've never known anyone who's worked with both Oracle and SQLServer who'd reccomend using SQLServer for enterprise work. Of course, the people with a financial interest in MS will disagree, but what else would you expect a salesman to say?

I've enjoyed reading all of the comments regarding the relative technical merits of MS vs. *nix. From a layman's perspective, here are some anecdotal experiences:

I've developed a couple of database Web sites that use forum presentation/search functions similar to FR, including pcaob.com (covering the new CPA professional oversight requirements), and karlnagel.com, my business consulting firm's website.

Each of the sites are hosted on a Linux/Apache server running MySQL and PHP. The cost is $10/mo for acccess to all the development tools, capped at 4 gigs/mo ie 100,000 page requests/mo x 40k/page. (Each additional 5 gigs is another $10/mo.)

The sites work fantastic. PHP is an excellent scripting language, and has numerous built in functions for PostGres & MySQL. I followed some of the style guidelines that JR established here to give the sites a really nice look & feel.

Everytime I review the activity logs, there are always a number of failed requests. When I checked with my Web host, he mentiond that they were MS .exe hack/virus scripts and led me through the details of the failure report. I don't even want to think about what could happen if these sites were running on an IIS/SQL system, much less the hosting & development costs.

I've developed simple financial apps in VB, and use Office/W2000 as my development platform for my PHP scripts (just using WordPad), so I'm not particularly religious about MS vs. *Nix. It's just for Net development activities, I can't see how MS can compete in the long-run with smaller companies (99% of all businesses) that elect to go with open-source.

MS has replaced IBM in the old adage of 'no one ever got fired for ..." in larger firms. If I was in charge of IT in a company with even more than 25 employees, I'd probably recommend just going with MS, especially if I didn't have a financial interest. In the long run, it's better to keep one's job than to be taking all sorts of technical risks.

But if I was involved with smaller businesses, or had a personal financial stake (as I do in my case), I would never just throw my money away at MS. *Nix is eating up this market, and savvy consultants selling into small businesses know it.

But if I was involved with smaller businesses, or had a personal financial stake (as I do in my case), I would never just throw my money away at MS.

Actually, there is a place for SQLServer, I think, in small businesses with simple needs and no in-house database knowledge. If your needs are few and your data is not really sensitive, and if you can deal with your site going down from time to time, then paying for SQLServer can be one good choice.

Of course if you ever grow, you'll need to migrate . . .

I'm a web developer for CSC, mostly internal project and resource management tools. 90% of my work is in Java, and I have to read and write to Oracle, SQLServer, Lotus Domino and in one case Access 97 (legacy DB no one wants to migrate, go figure).

I like MySQL. But I don't get to make the db choices. If I did, I'd use it for several of the projects I have. SQLServer's strength is in being relatively simple for inexperienced people to use. Of course, this can also be a weakness . . .

I view MS as the 'McDonalds' or 'Brittney Spears' of software. They make high-volume, low-quality goods that appeal to the masses. Now I eat McDonalds from time to time, just like I use MS products regularly. But if I were a professional caterer, I wouldn't use McDonalds food as my vendor!!!

I like the look and feel of your sites, very clean and usable. One of my biggest pet peeves with some web designers is they don't seem to realize that a web page is actually a piece of software and usability is the top requirement!

The better option, to me, sounds like to use Linux,...

You dork. The HP/Red Hat/Oracle solution isn't even available until next year!!! Meanwhile, you're going to have to use your right hand.

That's easy, the Oracle/Linux approach. With SQLServer/IIS/Windows you're locked into a single vendor for all products, and that vendor is known for stability problems. And expect to be forced into an upgrade path -- .NET -- that will lock you in even tighter to that single vendor. The one SQLServer db I have to deal with creates 80% of our DB problems.

What a joke. You call it "vendor lock-in". I call it "standardization".

With Oracle/Linux, you have a wide variety of choices of Linux vendors, a wide variety of web server choices, and Oracle is perhaps the most stable big DB around. The solution is more scalable without a doubt.

Of course, you can't get this combination until next year.

The single biggest thing to avoid as a businessperson is 'single vendor lockin'. Because then that vendor owns YOU. You're at the mercy of a single corporate entity. For example, if they decide the next version of their product -- like Longhorn -- isn't going to be backward compatible and you have to pay for all new copies of the same software, you don't have much choice.

Are you so daft that you're incapable of writing JDBC/ODBC/SQL stored procs that are vendor-independent?

As someone else has already pointed out, you get more processing power for your dollar with the HP/Win2K/SQL combination.

MS tried to buy phony high tpc numbers by using a $10+ million cluster!

MS used 3 times the hardware any other solution did, just to be at the top of the list. So clueless MS-only salesmen could push a lie.

As the numbers clearly show, on equal machines, the Linux/Oracle solution beats the Win2k/SQLServer solution in both price and performance.

And scalability, and stability.

Shoot, if you throw $10+ million worth of hardware at it, pretty much any software will scream. Even Lotus Domino would seem like a decent DB.

But machine for machine, Linux and Oracle win. Which I'm certain you know already, by the tone of your posts . . .

The HP/Red Hat/Oracle solution isn't even available until next year!!!

Let me get this straight -- You actually believe that Oracle on Linux and HP machines is not currently available???

You don't even know enough about this to know that the 'availability' in that chart is for that consulting company to provide you with the canned solution, and support for that solution?

You call it "vendor lock-in". I call it "standardization".

Like I call it 'copying the work of others' and you call it 'innovation'. We're all well versed in MS doublespeak.

Are you so daft that you're incapable of writing JDBC/ODBC/SQL stored procs that are vendor-independent?

Because of the way SQLServer does joins, they don't port to other DBs. I'm certain that's on purpose, to increase MS's only marketing tool -- monopolistic lock-in (excuse me, "standarization", in your language). And the 'conversion' tools don't work on anything even remotely complex.

MS tried to buy phony high tpc numbers by using a $10+ million cluster! MS used 3 times the hardware any other solution did, just to be at the top of the list. So clueless MS-only salesmen could push a lie.

Phony high tpc numbers? No, Harr. First of all, it's HP doing the testing. Second, if you need 709,000 tpmC, you're not going to get that from the Linux alternative. Third, the price is cheaper. You pay $39K per processor and get 709,000 tpmC. If you go with the Linux alternative, you pay $74K per processor and only get 138,362 tpmC. In other words, you get more bang for your buck with the HP/Win2K/SQL2000 option. That you can't reconcile this fact but want to instead call the numbers "phony" isn't surprising. You tend to demagogue things that contradict your vision of a MS-free universe.

As the numbers clearly show, on equal machines, the Linux/Oracle solution beats the Win2k/SQLServer solution in both price and performance.

We're not talking about equal machines here. Because there is no Linux equivalent running on this hardware. When there is, then you can come back and talk about it. Until then, you're full of shite.

And scalability, and stability.

There's no basis for this statement. This is pure FUD.

Because of the way SQLServer does joins, they don't port to other DBs.

That's a crock of shit. You can use pure SQL and MSSQL2000 will understand it. Obviously, if you use TSQL-specific extensions, those won't port; however, that is a choice, not a requirement. The fact that you can't figure out how to write standard SQL joins that work across DBs has more to with your own incompetence than any limitations in the product.

We're not talking about equal machines here.

MS cheated the test results, bought the top spot on the chart by spending $5+ million on more hardware than anyone else had and pretending that it was the database running faster.

Boy, you're awfully confused by all this.

I'll go real slow for you. Only one (1) Linux configuration has been tested. So you have two ways to do a direct comparison --

compare MS solutions that provide approx. the same tpc.
compare MS solutions on the same hardware.

In both cases, MS loses.

Look at the results, for goodness' sake.

Linux runs faster on the same machine. Linux can get the same performance on cheaper hardware. And it's certainly been my experience that Linux is more scalable and more stable.

And geez, you actually suggested that the cost per processor was relevant, without taking into account the differences in the processors???

Disclaimer: Opinions posted on Free Republic are those of the individual posters and do not necessarily represent the opinion of Free Republic or its management. All materials posted herein are protected by copyright law and the exemption for fair use of copyrighted works.