Slashdot Mirror
Linux Grabs World Record For TPC-H Benchmark
An Anonymous Coward writes: "Linux 2.4.3 now holds the world record by performance with IBM's DB2 in TPC-H. TPC-H is a decision support benchmark consisting of a suite of business oriented and ad-hoc queries and concurrent data modifications. This is way cool as the world record was held by SQL Server 2000 on Windows 2000 before." Caveats: this is only in the 100GB (smallest) category, and all but 2 of the other entries are several months old. Even so;)
>You can't honestly view benchmarks as: well, when Linux wins they are the holy grail, but when someone else wins, it's rigged.
Gosh, you'd almost think that rather than being a single entity which can only rationally hold a single opinion without being inconsistent that slashdot was a collection of individuals with differing opinions such that different subgroups can consistently hold differing opinions.
But hey, that doesn't get you the +5, so why not take the easy potshot?
And that's Microsoft problem:
Companies still need Unix (and unix experience) for the high end. The low end is eaten away by Linux. Their server might get crushed in the middle.
As much as SGI has some of the coolest stuff on the planet, they also are completely lacking in much of the needed vision and business competance to make their offerings work, and be profitable. They had until recently, a completely incompetant individual in charge of sales/marketing. They recently put a person in charge of day to day operations who may not be as committed to delivering value in the Linux market as you think.
The point is quite simple. They went for the fastest performance in a market benchmark that wants the lowest price performance metric. This benchmark provides a great deal of fodder for their competitors (NT) to point out that for small databases (this was quite small), NT is a superior price performance. Had SGI had a clue as to what was needed here, they would have worked the price performance side.
Raw performance matters far more at the high end of this market. The other issue which is hidden from view here is that the 2 TB file size limit in Linux is going to be an issue for the big databases. This issue will need to be addressed in the 2.4 time frame for the bioinformatics houses to continue to use Linux. The databases there are growing at an exponential rate, and are already in the 2-5 GB range, with TB sizes expected in the next year or two.
It is a shame that SGI did this the way they did, but then again, I am frankly amazed that they had anyone to work on it. I am even more amazed given the level of committment to IA32 (see the comments about followon products). This company needs to focus more efforts on potential money makers like IA32/Linux based systems.
just look at the other systems in that class. they run ~ $250,000 as opposed to the linux system which is ~ $1 million.
Um, did you notice that the linux benchmark was $347/QphH while the next one on the list (W2K) is $161/QphH. Which one is more expen$ive?
You may have to shell out thousands for the software (vs free Linux) but the _machine_ dominates by a lot. The Linux configuration was 4x the cost of the Windows configuration ($1M vs $250K).
There's really three different markets, with Linux winning two and Windows winning one.
Under 1 GB is dominated by Linux. It's cheap, it's fast. Hard to sell software to this market because the total budget is usually in the low thousands at most.
1 GB-100 GB is decent with Windows. Much more bang for the buck then the Linux solution. At the top end, Windows just can't handle the load and buckles. At the bottom end, it starts becoming unrealistic to really spend a lot of money on your database.
Over 100 GB is dominated by the hard-core *nices. Linux can probably be used seriously now although a lot of companies would rather go with the proven solution and pay. When $10k software is under 1% of your total purchase price, free OS vs paid OS isn't much of an issue.
Although it seems very logically that a 4 machine cluster is faster than it's single machine counterpart, it's not that simple an equation.
Not every database or operating system can scale that well. Lets take a look at each individual machine of the cluster. They are SGI 1450s with 4-way Xeon 700Mhz. The nearest competitor on the performance chart is the NEC 4-way machine, which is at 800. Assuming each individual machine of this cluster is also at about 800 tpm, then the cluster scaled at 85%. Not too shabby. Can you do the same type of scaling with Oracle? Not likely(look at any of the Oracle benchmarks, the biggest cluster they got have two machines).
Also, if you have recently read what the Oracle guys have been asking the kernel developers, you would know that there are a lot of features that the Linux kernel is missing right now causing it to not perform optimally as a DB server. The SGI and IBM guys have worked hard to get around every one of these barriers in order to get these results. This really shows that both these companies are very dedicated to make Linux be the top choice as DB servers in the future.
I don't want to burst anyone's bubble, but Linux off the shelve at the moment is still a very immature operating system for a DB server. However, with the work of companies like SGI and IBM, Linux now has a top result on a industry recognized benchmark.I wouldn't be surprise if there are more results coming in the near future.
Congrats to both companies for this great result!
This doesn't really do much for Linux, other than add fuel to the fire that "free software is more expensive". If you look at the results, Microsoft has an offering that costs $273k and gets 1700 QphH. The Linux offering, on the other hand, costs $950k and gets 2734 QphH. So for a little over triple the cost, you can get about a 50% increase in performance.
If you were really smart, you'd consider using Teradata on MP-RAS, which with a 1000gb database gets 18500 QphH, and only costs $700k, thus leaving $250k in your budget for end of year bonus checks.
--
"Don't trolls get tired?"
"But it's open source, so it's less secure. Would you entrust your data to something that anyone can modify the source code for?"
"But it's open source, so you don't have the satisfaction of having paid several thousand dollars for a Windows 2000 Datacenter site license."
"But it's open source, which we all know makes Baby Jesus cry."
"But it's open source, which sort of sounds like 'open sores', which is just gross, don't you think?"
Pick any or all of the above, submit to PR Newswire...
- A.P.
--
Forget Napster. Why not really break the law?
"Remember when the U.S. had a drug problem, and then we declared a War On Drugs, and now you can't buy drugs anymore?"
Err..what you just calculated is
/(Price/QphH) which is QphH^2 / price
QphH
which really isn't especially interesting.
You know damned well that behind the walls of corporations people will give each other high-fives when they strike a blow (even a percieved one) against their competitor. Why should the free software community be disallowed the same enjoyment of competition?
P.S. In your face, Microsoft.
The "cue the foo posts in 3, 2, 1..." posts will commence with no subsequent foo posts in 3, 2, 1...
hawk
> unstable and require a complete reboot
See, a perfect example of the verb "to Gates", as explained above
hawk
> little part does.
uhm, no. At least not in the past. Part of their problems with DOJ have been about *not* disclosing such information to competitors to give word/excel a leg up, delayed disclosure, and flat out mis-disclosure.
Word and excel can rely on undocumented materials. Anywon else faces the probability that a change will break their code.
>. The only hold microsoft has over
> anyone else is their research budget!
THat and the windows monopoly that let them strike the licensicing deals to include Office on most computers. This is the single largest factor in the displacement of Word Perfect and Lotus by Word and Excel . . .
The reason that people hate Gates is that they've been hurt by his software. As in, "I've been Gatesed," a reference to the loss of data, work, or effeot because a machine behaves in a fundametnally unreasonable way, or becuase the command just used is different thatn the command that was bound to the same keys on the prior version, resulting in a loss of data. Or when the machine refuses to allow acces to a file without a bizarre workaround. Or when focus to a window is lost even though you were still typing because of a "helpful" feature. Not to mention the times when this week's version of word/excel has problems with a vile saved by last week's.
No, it isn't his money, wealth, or succes that cause most people to hate him. It's the painful to use products.
I never hated windows until I had to spend a day with it to et the files I needed to download Linux . . .
And no, I don't hate Gates. I'm not even anti-microsoft. I do, tough, understand someone who regularly uses the products becoming so . . . and I have the advantage that I haven't spent a lot of time with any since word 5.1 and excel 4--not from loack of opportunity, but because the later versions took out features I used constantly. I switched to *nix over LyX, not ideology (and I would have had to switch in a year or two anyway due to high performance computing needs that windows just can't handle).
hawk
Oracle actually supports both actually (java & C). The current versions of Oracle support java inside the database engine itself as well as the execution of external code built in whatever language you like.
You're not limited to PL/1 with embedded SQL.
Even so, the real usefulness of java inside of Oracle is it's [java's] standard libraries rather that the language itself.
A Pirate and a Puritan look the same on a balance sheet.
"safer" means that systems crash less and you actually have a real choice if you decide that your current solution is unsuitable or overpriced.
"safer" doesn't mean that you get a free lunch.
A Pirate and a Puritan look the same on a balance sheet.
The two positions are not at all at odds with each other. One can still point out that you can be beat at your own game while still claiming that your entire game is bogus.
Also, there were some real issues raised with the manner in which Microsoft based products obtained their "high scores" in the TPC-C.
The Microsoft solution used what was essentially 100 discrete databases when compared to single IBM or Solaris instances.
The same was true of Mindcraft.
Hell, the same might even be true of this particular SGI beast.
Either way, Microsoft's "damned lies" can't be allowed to exist in a vaccum.
A Pirate and a Puritan look the same on a balance sheet.
Actually the typical manager will look at the price tag first, realize that the Linux solution costs 4 times as much for only a 50% gain in performance...
He will then buy the Microsoft solution which offers more than enough performance for a signifigantly smaller hit to the pocketbook.
Just a note: Although the database was IBM's DB2, the benchmark was done by SGI on 16-way Pentium III Xeon cluster. Also of note is the fact that the nearest competitors were 8-way Pentium III Xeon systems, not clusters, running SQL Server 2000 on Win2k.
Unlike the TPC-C benchmark TPC-H is a real world measure. As such this is a useful, if limited, result.
Maybe nobody on Slashdot needs a 3TB database. And on this benchmark your bank (which will have a 3TB DB) won't yet be taking up Linux. But it is a step on the way.
That is probably coming in the future. The reason logs are capped at 4G (or 32G, depending on the version you're running) is likely because back when DB2 was designed, disks were small and expensive and admins would want a way to limit the amount of disk space consumed by the logs. In any case, if you have a transaction that has not completed in the time it takes for 32G worth of other transactions to occur, you may have a design flaw with your application.
it's too easy to get into a state of questionable integrity, at which point DB2 stops. Just run a LOAD statement and leave out the word 'UNRECOVERABLE' - bam! Tablespace is in backup pending state and you can't access the data.
That behaviour is there to preserve data integrity. Loads aren't logged, so you will have to create a backup image that contains your loaded data. Either specify COPY YES on the LOAD command so you'll get a load image that can be applied at rollforward time, or back up the tablespace you loaded data into. Otherwise you will end up with data that can't be recovered in the event of a disk crash/building burning down/etc.
On-line backups have only just been added to the Unix version of DB2
Online backups have been in DB2 for years. Perhaps you mean to say "incremental backups"?
IBM *did* post a Linux score. What do you think this entire thread is about?
So now Linux doesn't exist as a desktop OS? Surely the 1,000 posts every hour praising the penguin prove differently?
Is it impossible for both Linux to exist as a desktop OS and Microsoft to have monopoly power in the Desktop OS market?
The real monopoly power lies in Microsoft Office, and the fact that it runs on Microsoft OSes (plus the insignificant MacOS which has been relegated to its comfortable and harmless niche).
This allows Microsoft to spead its Desktop monopoly power into the server space. WinNT works better with all the desktops (which are, of course, all MS, thanks to Office), so therefore the demand for WinNT grows.
Is it becoming clearer now?
Until Linux can run MS office faster, easier, more effeciently, with less problems, and an easier installation than a MS OS, then Linux will not be a threat to MS in the dekstop market.
But there is a little bit of merit to your needed accusation that Linux supporters are two-faced. And I'm saying that as a die-hard user and supporter of Linux.
I don't make the rules. I just make fun of them.
The reason is that Oracle, by default, has been using rule-based optimizations for quite some time. It has cost-based optimizations, but they aren't turned on by default (actually, I think they are in 8i). Noone uses them because their queries are now so optimized for rule-based, that switching to cost-based would make it quite unpredictable. Oracle Applications 11.0.3 and earlier is made/optimized for rule-based queries. For any given query, you can turn on cost-based optimizations by adding a specially-formatted comment (forget what this is at the moment). Anyway, Oracle Apps 11i w/ Oracle 8i is supposed to have cost-based optimizations by default. Now, rule-based optimization is more tradition than anything.
Please note that much of this comment is hearsay from what the DB guys are saying. I try to stay out of most of the Oracle stuff myself, unless I have to mess with it.
Anyway, I'm curious whose cost-based optimizer is best.
Engineering and the Ultimate
(SQL Server 2000 is also *very* robust. FAR more robust than any free solution, and easily ranks up their with DB2 or Oracle.)
:p). I'm disappointed that more SGI systems have not been benched: SGI memory architecture is really tits, and if you're going to pay that much for RAM you should at least show it off.. And how about the 5xx RS6ks with the interleaved memory planars? Or am I dating myself here ;)
You can thank Sybase for that.
Also, keep in mind that SS2k runs out of gas at 300GB bases, due to scalability issues with its underlying OS. Linux 2.4 currently has similar limitations (shown by the SGI system getting only about 40% more performance with 2x the CPUs
Don't forget to use the right tool for the job. Linux has come very far, but it is definitely not ready for performant terabyte DBs yet..
Your Working Boy,
- Otis (GAIM: OtisWild)
he he he
They're offering the challenge to web app systems (DB + J2EE).. You'd need to find a really fast J2EE solution for Linux (is WebLogic ported?), and even then the contest T's and C's are kinda harsh (we get to tune your production system for 90 days is one...)
Your Working Boy,
- Otis (GAIM: OtisWild)
Anyone know how well the free transactional RDBMSes (Postgres and sapdb) do on these benchmarks? Do they at least manage to complete the run?
-- Ed Avis ed@membled.com
There's one thing people forget about the test: the SGI machine ran 16 CPU's, while the other machines ran 8 CPU's.
I want to see what the benchmarks look like on the SGI machine with 8 CPU's for a much more valid and fair comparison.
Raymond in Mountain View, CA
I want them to repeat the test with the SGI machine running only 8 CPU's for a fair comparison.
My guess is that with only 8 CPU's the DB2 benchmark results may not be faster than what was achieved with Windows 2000 Datacenter on 8 CPU's.
Raymond in Mountain View, CA
For the most part the articles are about what Rob & co. are interested in, but have you looked at some of the questions in Ask Slashdot recently? It seems like Cliff will post any question that hasn't been posted before, or has but asked in a slightly different way. Also, there have been many times that an article gets posted just so people stop submitting it, even though they'd rather not.
It wasn't Slashdot that questioned TPC's worth, it was KhaosSpawn. Just because they post an article doesn't mean they necessarily agree with it.
For all the talk of Linux vs Win2k, when I look at the benchmarks, what I notice immediately is that IBM's DB2 is at the top of most of those lists, and Oracle is not. I know that Oracle is best known for handling extrememly large databases, but isn't it supposed to perform well too? It only shows up once in the list, where the Sun/Oracle box beaten *badly* by a less expensive Intel box!!
Anyway, the reason that struck me is that I'm very interested in DB2. I've heard good things about it, but not nearly as often as I hear about MS-SQL, Oracle, PostgreSQL, and MySQL, all of which are covered in O'Reily's "SQL in a Nutshell". I was only disappointed that this book didn't cover any others. If I wanted to get similar information for DB2, where would I look? It'd be cool to see some DB2-lovin hippies to publish information in the form of an addendum to that book for comparison's (and reference's) sake.
You've made the classic mistake of confusing a vocal minority with an important majority or even an important minority.
/. readership which is posting in favor of taking these benchmarks seriously.
/.
No sane person debunks benchmarks and then considers them seriously meaningful, and most linux users are sane. Remember that Taco & crew are editors trying to push page-hits up, they're not a real part of the linux community. Just look at how much Taco runs linux for to see that.
The Slashdot editors stopped being a real part of the linux community a long time ago, mostly when they turned commercial. They have some pro-linux sentments largely because, I suspect, the crowd that would be pleased by this is a decent portion of the slashdot viewership.
As well, people who like to yell and scream whenever they see linux partisanship start generating more page counts when Taco & crew starts getting into the irrational linux-mentioning stories.
As Hanlon's razor goes: "Never attribute to malice that which is adequately explained by stupidity." Just substitute greed (or more charitably desperation) for stupidity, and the current situation has a very good explanation.
If you doubt this, just take a look at the percentage of the
It's always dangerous to judge the majority by a vocal minority, and especially so on
They laughed at Einstein. They laughed at the Wright Brothers. But they also laughed at Bozo the Clown. -- C. Sagan
There are sometimes in the course of human events institutions which must be destoryed, institutions whose very existence is an impediment to better things.
Microsoft is out to take over the world. They want to own the information architecture to the world, and they have repeatedly demonstrated that they will use almost any tactics necessary to do so (to my knowledge they haven't had anyone killed yet).
Microsoft is out to conquer the world and control it. Microsoft must be destroyed so that this is not possible. Every time that microsoft gets hurt, it is theoretically one step closer to being destroyed.
This is what you percieve as hate - many people want to live free lives where they are not controlled in most details of their life by some one person.
This is all very simple and has been said over and over. The fact that you don't understand it seems to be strong evidence that you're just a troll.
Consider the evidence:
1. Microsoft wants their products to be everywhere
2. Microsoft has demonstrated that they will compete using all legal techniques, not just on merit (vendor lock-in and monopoly pressure being their two favorites)
3. The world is moving more and more towards being information centered. Look at the various "intelligent" devices which are being created and more are being slated. Look at music that will require special devices to play, etc. Just think of when cars are networked and require a valid windows license (revokable at microsoft's whim) to operate.
Microsoft is bad for the world. In general most free people want a democracy or something equivalent, they don't want the world controlled by bill gates.
Which part of this do you not understand?
They laughed at Einstein. They laughed at the Wright Brothers. But they also laughed at Bozo the Clown. -- C. Sagan
Very enlightening - I stand corrected.
Caution: contents may be quarrelsome and meticulous!
Your right to not believe: Americans United for Separation of Church and
...it's more the fact that Windows isn't the leader in any other category (all the others appear to be heavy-duty *nix). Although I imagine Microsoft will be coming right back with some results of their own. Thanks a bunch to whomever submitted this configuration for testing - does anyone know if it was sgi, or who?
Caution: contents may be quarrelsome and meticulous!
Your right to not believe: Americans United for Separation of Church and
Maybe now Enterprises will realize that linux is a big mans toy, but actually, is quite a powerfull Operating system, that can compete with the expen$ive OS's out there... Hopefully more support from other Vendors for linux will result... Go Linux Go
I'm not jealous of Microsoft's success at all. I would like them not to have their current 'success' because they abuse it to ensure things stay that way. What I would really like to see is a large number of smaller companies sharing in a great success, with none of them dominating the rest (as is the case of Microsoft in the 'Windows market'). Open standards and open source, along with a change in business models could bring about this kind of situation. The current situation is that the mantra of every other industry that requires consumer participation, 'our business is driven by our customers', which seems a fair proposition, just doesn't exist in the Microsoft-dominated world, where it is standard practise to have a clause in the license that if the product doesn't work or causes you damage in any way, that is in no way the fault of the supplier, and you have no recourse.
People hate Gates because he is litterally out to destroy Linux. If he could find a way to obliterate Linux from the face of the planet, he would do it, along with everything else that is non-Windows. In contrast, most Linux users prefer a heterogenous software world, which includes Windows along with everyone else. Microsoft constantly try to escalate the situation though, by trying to break open standards for example, or making speeches about how the government ought to do something about free software.
Essentially, it isn't safe from a free software perspective, to just think of Microsoft in the same way as any other large closed source software company. Other large closed source software companies don't go out of their way to attack free software, they try to provide non-free software that stands on its own merits. Turning your back on Microsoft is like turning your back on an axe-wielding homicidal lunatic.
Well, the 4 machines SGI used cost around US$450,000. It is certainly reasonable to ask if a benchmark performed on that sort of hardware has any relevance to 99% of users. And they used RedHat 6.2 so it's hardly cutting-edge Linux technology.
Just for giggles I priced four Penguin Magnus 4500's (as closely configured as I could to the SGI boxes). They were just over 1/4 the price of SGI's hardware.
It is also worth noting that the DB2 EEE licenses were over 1/3 the total price.
I see it as interesting because previously no vendor had used Linux for a TPC benchmark (indeed I was under the impression that the TPC wouldn't allow it).
TPC doesn't disallow anyone from running TPC benchmarks on Linux. They disallow anyone from publishing TPC benchmarks that isn't a member of TPC and that aren't officially submitted to TPC. Both membership in TPC and submitting results cost money. Up until now a vendor hasn't been willing to pony up to submit TPC numbers under Linux. That is apparently changing.
I'd really love to see IBM and/or Compaq submit DB2 EEE results on a cluster of four quad Xeon Netfinity or Proliant servers... I think that IBM or Compaq could offer a significantly better $/QphH value than SGI does, because SGI's hardware is expensive.
I mean come on, they were even using IBMs DB2 database, are you telling me that IBM doesn't know how to use their database in a cluster of machines but SGI (RIP) does?
SGI was just the first to submit a Linux based TPC entry. SGI started working seriously with Linux before IBM had really taken that up too. It will be interesting to see if IBM follows SGI's lead and submits a similarly configured Netfinity (x Series) machine with DB2 EEE. I'd for one love to see that.
Is dynix more important to them than Linux, it looks like it to me.
More important to IBM or more established at IBM? Linux doesn't (yet) run on the NUMA/Q hardware that DYNIX/ptx does. However, look at what IBM is doing with AIX 5L. They are starting to make it more and more like Linux, and make Linux more and more like AIX by releasing key parts of AIX technology as open source for Linux. IBM has previously stated their intention not only to get Linux running on all their hardware platforms including the NUMA/Q machines, but also to bring the technology for DYNIX/ptx to Linux, and one would assume also add "Linux Affinity" to DYNIX/ptx.
Actually, Data Center only does up to 32 CPUs (http://www.microsoft.com/windows2000/datacenter/e valuation/business/overview/default.asp). However, that's still pretty impressive for x86 hardware.
How much does DB2 for Linux cost? I've seen "free beer" DB2 CDs, but what if you want to use it for a business? Have you had an stability problems with DB2 on Linux? I've heard that Oracle on Linux can be moody.
cpeterso
The price per transaction is more than twice as high. Sure, you can pay a billion dollars and get great perf. Good ol' cheap Linux :-)
I've seen a lot of posts in this thread knocking Linux's victory here because it doesn't have the best price/performance in the category.
It's my understanding that TPC-H is NOT a price/performance contest. It is a performance/1 contest!
Does TPC have a price/performance category? If so, let's look at the results of that category to argue price/performance.
TPC-H is about performance.
Please, somebody correct me if I'm mistaken about this.
Joe Mainusch http://www.weber-amps.com
Yeah, because we all know that the Linux Corporation has lots of extra money to throw around on bought-and-paid-for benchmarking tests. Companies with real integrity (like Microsoft) would never do something this transparent!
Joe Mainusch http://www.weber-amps.com
9iAS is a fantastic product that can't be beat.
But what is the point, 9iAS is available for Linux. Kind of like saying your going to prove Linux is better then Oracle when the Oracle solution on the same platform will outdo DB2 by a longshot.
I can gurantee that one. Give me Oracle 9iAS on the same SGI hardware and it will smoke the DB2 machine on a true dynamic benchmark as well as static benchmark.
Or you can also wait a few weeks for Oracle 9i for a real suprise :)
2. That's probably true. Linux support for hot-plug SCSI is weak right now. Don't expect it to stay that way for long. It isn't exactly rocket science to develop a driver to talk to the SCSI hot-plug controller :-)
3. Actually, it wasn't just OS cost in this case. Check the database software cost. It's pretty significant. I still think you're right and that Win2k will win on a single 4-way at the moment, but the scalability of the clustered solution is impressive. Of course, I suspect this has a lot to do with DB2 EEE edition, since it is highly optimized for this kind of setup (think IBM SP2).
Tim
now you're on the trolley!
[We don't come from a planet. We come from a grid sector.]
Also, there have been many times that an article gets posted just so people stop submitting it, even though they'd rather not.
:)
Maybe we need a new slashdot topic "Enough Already!!!" that we can uncheck if we don't want to see articles that everyone submits but they don't want to post. Then just make it a slashbox so people might be forced to see the top X headlines, but at least it will go by fast
This space for rent. All reasonable inquiries will be entertained at proprietors discretion.
The SGI and IBM guys have worked hard to get around every one of these barriers in order to get these results. This really shows that both these companies are very dedicated to make Linux be the top choice as DB servers in the future.
You're right, they are very dedicated. The reason is they share a common enemy, "Microsoft". Without Linux as an OS SGI machines are going to dissapear as mid-high level workstations (MS Sure won't support them), and if they can be freed to concentrate on Hardware over Software, their server line might keep selling (they have a great product, but to quote an engineer I worked with 3 years ago when expecting a Solaris machine and logging into an SGI machine "Irix? What is Irix?").
IBM's grudge? Gee, can we say OS/2? Okay, so they probably don't still hold a grudge over that, but they have certainly bucked horns against MS every chance they could (or so it seems).
Its great to have two heavy hitters in our corner. Lets just remember that this is still business and people are still greedy. They're here because it suits them, which suits me fine. Common goals can take you preaty far.
This space for rent. All reasonable inquiries will be entertained at proprietors discretion.
If you want 2733 QphH then it is the only contender.
On a unit cost basis it would also be interesting to see single CPU results.
It's odd to hear carping about Linux being tested on an unfairly large machine!
... since when did doubling the amount of performance only cost twice as much money?
Never? Have you ever seen anything like it? Even something as simple as getting double clock rate from Intel costs about 4 times the amount of money. Is someone somehow not expecting that going from extreme performance, to twice the performance would NOT cost 4 times as much?
1. Doubling the amount of CPUs does not yield double performance, but cost much more than twice the amount. This would happen to the Windows 2000 servers as well.
2. Did anyone at least check out what is included in the price? I do not know, but for all I know, the SGI 1450 may be equipped with more levels of redundancy, have more expensive (more stable, not fastedr) hardware.
3. The operating system cost for one of these beasts is miniscule, so I as a Linux-advocate would not even consider arguing for price of the OS for machines like this. (I'd argue performance, stability and tweakability).
I urge you to go and look at the LKML, or, more likely, KT. Specifically, the thread about X15 compared to TUX (the first is user space, the second is the kernel web _accelerator_ you mentioned).
X15 is faster. Not much, but slightly.
It's an issue of design, thorwing it into the kernel does not, inherently, make it faster, in general.
--
Almost any medium-sized retail company with a reasonable merchandising system is going to be looking at a DB in that size range. When you start tracking every POS transaction for a couple of years and doing analysis down to SKU level, things add up.
We would, but only because MS has made it very clear that this a battle to the death.
If Linux was produced by 1 company, it would be doing just as unscrupulous acts as Microsoft, but all you Linux lovers would be up in arms defending Linux.
First of all, that's impossible, so your point is moot. But more importantly, you're wrong: did you miss all of the RedHat flamefests here? RedHat is far from the only company that produces a distribution of Linux, but they are repeatedly flamed (and worse) when they try heavy-handed tactics.
To within half a percent, pi seconds is a nanocentury. -- Tom Duff
Exactly. The average PHB who sees/hears of this story will only remember the headline: "Linux Grabs World Record For TPC-H Benchmark". They're not going to read the results, look at the configuration of the boxes, and compare performance per dollar, etc. All they'll remember is that Linux won.
Sure, the headline is misleading propaganda, but as we all well know, propaganda and marketing are what sells products, not technical superiority. Because of this result and others, Linux is now seen as a viable DB competitor. And that's a good thing (for everybody).
To within half a percent, pi seconds is a nanocentury. -- Tom Duff
What's wrong with you?
Why does it seem that most people spend most of their time hoping for something bad to happen to Microsoft? If there's a security problem with a Microsoft product, it's not a *good* thing, much less is it "too good to be true"! If Linux scores better in a database test, it's a good thing because it means there have been advancements in software. It's not a good thing because Microsoft came 2nd.
The whole culture of hate here on Slashdot (and in the open source community in general) really bothers me.. Why do you have to hate something? Why isn't it enough for you that Linux or your favorite open source project is successful and works great? Why do you have to stomp everything else? No wonder people say "open source, closed minds".
If it was just 13 year olds writing the comments, I'd understand it. But it's the editors of Slashdot too! You guys really should set a better example than that. Even Linus Torvalds said in a recent CNET interview that he doesn't understand why everyone hates Bill Gates so much.
It's much more productive - and much better for the cause (which is to make better software, remember?) - to focus your energy in positive things. Write software, report bugs, test.. Sure, celebrate Linux being first in a database test, but don't celebrate it because it knocked away Windows 2000 & SQL Server 2000 from that spot.
Define yourself by what you are for - not by what you are against.
As long as the good does not fight back the evil will win. MS is well funded, well armed, and very experienced at destroying competition. Sitting idly and saying we are the good guys will result in MS killing us all.
War is necrophilia.
Hate is blowing up a federal building and killing hundreds, hate is firing live bullets into a crowd of protesters, hate is carpet bombing cambodia.
Hate is not posting on slashdot.
When Bill Gates dies at the hand of a user subjected to a lifetime of windows that will be hate.
War is necrophilia.
When they buy a toyota they get a choice of engines, more then that they can choose not to buy a toyota at all. Besides Toyotas are exceptional products which work as advertised and have guarantees that assure the buyer that if anything goes wrong they will get fixed for free.
War is necrophilia.
It's hate nevertheless. Even more so because the measure of civility of a nation can be measured by how it handles dissention and protest. The measure of a police officer or a soldier is how he or she reacts to stressful situations.
War is necrophilia.
I disagree. MS has a history of destroying technically superior products with lies, threats, extortion, and rackateering. If you fight MS by "only" making a superior product MS will crush you.
You have to fight back with everything you have. Your brain, your mouth, your money anything less then 100% commitment will mean the end of linux. MS has billions of dollars in the bank, hundreds of politicians in their pocket, thousands of lawyers on the payroll. They will crush us like so many cockroaches.
As for your hardware comment it's plain weird. They got twice the performance and spent twice as much. This is practically unheard of. Ususally you have to spend exponential amounts of money to get linear improvements in speed. They can be proud of their accomplishments.
War is necrophilia.
"The business world i live in has ordered, civil, "real-world" conversations regarding our software and development platform - and the tactics you espouse only damage the cause of GNU/Linux and OSS."
The business world is far from civilized. It's full of duplicity, lies and cheating. I know because I have run businesses and of course worked in them from being a serf to being in management. To paraphrase George Carlin. You know business people are assholes because if you ever put two across the table from each other they will both be convinced the other guy is out to screw them.
For some reason it does not bother your boss when Mr. Allchin calls us communists, it does not bother him when the top MS staff talks of "knifing the baby" and "cutting off the air supply", or when MS executives lie under oath or tamper with evidence during a federal trial. It certainly does not bother your boss when Mr. Ballmer tells lies in public. He is after all a PHB.
His eyes are glassed over by the shiny ads he sees in businessweek. All we can do is point out that these people are liars and have no moral or ethical compass, we can point out they routinely screw their partners over and hope that it will reach one or two CIOs. It won't effect most of them but honestly we can't afford to fly them out and feed them at fancy restaurants so it's all we can do.
War is necrophilia.
PostgreSQL has its own PL/SQL-like language built in. You can also use Perl and (iirc) Python and TCL as extension languages.
Become a FSF associate member before the low #s are used
1. I checked out Compaq's site and there is about a $10,000 difference between a 4 way system with the Profusion chipset and without it. That's about the cost of 2 CPUs, so you aren't exactly correct here but my initial point wasn't exactly right either.
2. True enough. The Linux machine also included the cost of the UPS which wasn't part of the Compaq system.
3. Interestingly enough, the cost of software on the Linux system (including maintenance) was actually $365,000 where the Windows system was only $27,000... So much for MS being expensive!!
Fear: When you see B8 00 4C CD 21 and know what it means
If you compare the Linux solution to the next Win2k solution, check out the software prices:
Linux + DB2 = $366,000
Win2k + SQL2000 = $27,000
OMG!!! No wonder Microsoft wins sales in the database arena!
Fear: When you see B8 00 4C CD 21 and know what it means
So basically you are saying that 4 linux machines can beat one Win2k machine. In fact, Linux is close to half the price/performance of any of the machines there.
What it comes down to is that if you want to save money then Win2k is going to give you the best performance for a given price, up to 8CPUs. As there are no benchmarks for a Win2k cluster in TPC-H, you can't draw any conclusions from this 'win'.
If you look in TPC-C however, you'll see that a Win2kAS clusters are ramming Unix right up the hole in performance and price... I'd be sticking to the 'Do TPC benchmarks really mean anything' stories if you want to promote Unix over Windows.
Fear: When you see B8 00 4C CD 21 and know what it means
No matter how SGI got there score, where is
the so called Linux company IBM?
I mean come on, they were even using IBMs DB2
database, are you telling me that IBM doesn't know how to use their database in a cluster of machines but SGI (RIP) does?
I think IBM *does not* want to post a linux score while they can still sell their proprietry machines like as400 and the like. Look at
the OS they have submitted, AIX, dynix?!? and win2000. Is dynix more important to them than
Linux, it looks like it to me.
Why is it that so many sites like this one offer spreadsheets for download as "XLS" when in fact the contents of the file you receive are simply plain text? Just because PC users are too dim to load Excel themselves?
No wonder so many people think they need Office to make effective use of the web. How about if I start making all the images on my site "Gimp" images when they're really just PNGs?
Hmm. actually that's not much more mean that what I've done to my homepage. Try that in MS Win IE - oops, someone doesn't understand W3C font/charset interaction recommendations.
Mea Culpa, apparently this is not a single machine...
Don't sweat the petty things. But do pet the sweaty things.
the SGI 1450 Server as 16 procs
2733/16 = 170(TCP) per CPU
Um, the SGI's page for the 1450 states that it's only a 4-way machine.
That yields 683.25 TPC/CPU for Linux, versus 212.375 TPC/CPU for win2k on the Compaq hardware. In other words, Linux is more than three times faster according to my calculations.
One of us has our numbers wrong... I've got SGI's website to back mine up... Am I missing something here?
Regarding w2k datacenter, it only supports up to 32 CPUs. Anybody know of any shipping x86 hardware at that scale? Last I heard Unisys had the only box, to be resold by Compaq... Until Compaq backed out (don't mean to troll here, but conspiracies abound that MS pushed it that way in light of w2k dc's "shortcomings" - YMMV).
Don't sweat the petty things. But do pet the sweaty things.
The price/performance is what is of concern here, and SQL Server 2000 with Windows 2000 still comes out far ahead of all the rest.
:-p
Is this a good or a bad thing?
Well, if you're not a moron, it's a good thing since you can have the best performance for the smallest buck. (SQL Server 2000 is also *very* robust. FAR more robust than any free solution, and easily ranks up their with DB2 or Oracle.)
If you're a moron, however, please begin crying now. Use what's best for the job DAMNIT.
Microsoft has hired Al Gore as a consultant while they plan on having TPC recount the benchmarking process which they labeled as unfair.
"This is ludicrous, no way are those results in any way proof of anything more, and with a 52 billion dollar budget allocated to marketing we'll prove it." stated an anonymous Microsoft spokesperson.
According to an insider, MS is also planning on declaring war on SGI for not using their ZX-10 servers for the test. An SGI spokesperson simply states, "We tried to use the fucking server, but the shit just bluescreened before we even turned the son of bitch on, so we kicked it a few times and chucked that shitty OS for the benchmarking test. Look, in all fairness we tried, but until Microsoft gets their shit together, we're going to try to capitalize on using a free OS such as Linux. Our company isn't do so great anyways so Microsoft can take their licensing and shove it."
After conceding to SGI, Microsoft then contacted Mr. Gore in efforts to gain insight on what not to do when asking for a recount. "We called Al to see where he screwed up asking for a recount so we don't fall victim to the same outcomes. We then tried to have him sell XP licenses at Columbia, but he conceded when pushed too hard." stated the anonymous MS employee.
Stay tuned as things get ugly
AO -- removing the dot in dot com
Want Root?
The SGI box is described as being clustered. So presumably that means a 4-way cluster with 4 CPUs each, versus the Compaq's single-node 8 CPUs. Which means the Linux option still costs way more to implement. Back to square one!
"Life is like a sewer - what you get out of it depends on what you put into it" - Tom Lehrer
QUOTE: If you justify hype like this, then you are no better than MS's FUD teams And you didn't know this already ?
Linux is on top there too, at $347. The rest is way cheaper at $161 (second place) till $267. This tells me that either the hardware from SGI is too expensive, or Linux isn't using the available hardware efficiently.
You made me think. What is going to happen to databases? I personally, predict an explosive growth, and here is why:
1. More and more people access the Internet and various services there
2. Users are lured by providers who offer richer services
3. The world population is growing geometrically (duh)
From this point of view, it would appear to me that the high-end database business is poised to be lucrative. Linux is still not there, but it will, quite soon. But don't expect Solaris to stay still. Solaris is evolving, too, and now their (Sun's) hardware is getting quite cheap (unlike HP's L,N or V, or IBM's RS/6000).
So, I think Linux developers should seriously target the large db business, but by no means should this quality leap be done sloppily.
Sigged!
Still, we could try to slashdot Walmart's stores, nationwide. Imagine, if at exactly 10:33, suddenly every store started reporting purchases of a bottle of shampoo and a bag of chips!
--TheOrangeSquid Is it any wonder things seem so awry? We swim in a sea of confusion and don't have to think to survive
3 TB databases are large, but not unusual.
Most large retailers would have databases this big.
A few years ago I read that Wal*mart had 101 TB.
Most of that is in a Teradata database.
I was one of the original developers of the Teradata database system. Man, do I miss it. For big decision support systems, nothing else on the market comes close. It's much easier to administer, and is the ultimate in scalable.
It almost died when NCR/AT&T bought it (they had no idea what it really was), but nowadays NCR has finally recognized it's worth.
Many companies with REALLY LARGE database rely on Teradata. For example, a few years ago I read Wal*Mart had 101 TB in their database (they were always Teradata's biggest customer).
Read the full disclosure report. They used four machines.
Walmart, the world's biggest data center (other than maybe the NSA, etc...), is about 50 TB, running on Oracle on Solaris, I believe.
There's 10 types of people in this world, those who understand binary and those who don't.
Way back in 99, there was an article in ask slashdot about it http://slashdot.org/askslashdot/99/02/17/1813211.s html.
Seem to remember it mostly being video cards hogging the bus so they can win at benchmarks.
The "culture of hate" didn't just spring up overnight out of nothing. I've hated Microsoft when I didn't know a single other person who cared about the issue. I've hated Microsoft before I knew of any alternatives. I hated them for making computers ugly and unusable. I left the computer world to escape Windows. With the rise of free Unix I reentered the computer world.
Hate is the gut feeling I get when I have to use a Microsoft product. I didn't get my hate from slashdot or anywhere else except Microsoft's deliberately vile products. MS is a creeping pestilence and the sooner it's eradicated the better.
By the way, I don't care what Linus says. He doesn't speak for me or for any Linux users. He can go to work for Microsoft and my feelings towards Microsoft won't be any friendlier.
The linux entry also used twice as many processors as the second-place SQL Server 2000 running on Windows 2000. It was also not competitive on a price/performance basis.
/. users hate Windows or think Microsoft is out to get them!
They are also using four machines rather than a single machine -- which means four times as much to manage -- as well as twice as many processors. MS SQL Server is also 25 percent more efficient than them for CPU.
So can linux actually claim this a win? There are 4 machines compared to 1, and it costs over twice as much for less than twice as much performance. I can't IMAGINE how you kids would react if MS put a cluster against a single linux machine in a benchmark.
____________________
Remember, not all
Prevent linux based DDOS's!
http://linux.denialofservice.org/
I think the OP meant that Linux is the youngest contestant. NT was released after Kernel 1.0, but development started well in the 80s.
> The Linux kernel is more than 10 years old and it still
> has a toy-VM, toy-filesystem(ext2fs) and toy-scheduler.
A toy compared to what? BSD? Commercial Unices?
Anyway, there is a lot of activity regarding the file system issue. reiserfs is part of the official kernel since 2.4.2. sgi released version 1.0 of their port of xfs port. And IBM is still working on jfs.
Hardware vendors have discovered that Linux can be used to outsource OS development.
--
Post tenebras lux. Post fenestras tux.
at least per CPU
the SGI 1450 Server as 16 procs
2733/16 = 170(TCP) per CPU
ProLiant 8000-X700-8P as 8 procs
1699/8 = 212 per CPU
and doesn't the DataCenter(tm) version support like 64 procs?
Anyway, i think it's pretty damm cool linux is on top, it reassures me that TPC.org isn't just one of Microsoft bit-achs.
-Jon
this is my sig.
TPC-C is the reference benchmark. Linux has no entry there. http://www.tpc.org/tpcc/results/tpcc_perf_results. asp
Also, this is a measure of the Database software speed. DB2 and SQL2000 are quite fast and on any 32-bit x86 OS they would beat the pants off of mySQL or some of those other freebies.
Look at the TPC-C benchmark to see the real battle between Microsoft and Sun.
Okay, so a few months ago, you ran an article: Are TPC Benchmarks A Worthwhile Measure? where this test was derided as being a worthless measurement? It was seen as "not realistic" because nobody needs those kind of servers... At least on Slashdot.
So now that SGI cranks out a server with twice the processors and knocks off a half year old record, it's legitimate because Linux wins?
This is absurd. Either this is a legit benchmark or not, make up your mind. If you justify hype like this, then you are no better than MS's FUD teams.
You can't honestly view benchmarks as: well, when Linux wins they are the holy grail, but when someone else wins, it's rigged.
Alex
Maybe I'm missing the point, or maybe I'm not clearly stating mine. DB2 does (and has done since the early 1980s) statistics-based optimization of queries *automatically*. Every time I talk to Oracle developers and DBAs they tell me things that sound like sheer insanity to me: what order you list clauses in your 'WHERE' dramatically change performance, you should do an EXPLAIN PLAN and then write your query accordingly.
I agree that the best optimizer is the human mind, but I have yet to meet database programmers who can bring themselves to regularly remember to care how big the table they are accessing is, or what fields are indexed or which part of their cartesian product is bigger. Programmers don't care. Oracle programmers have to (as far as I can understand). DB2 programmers don't.
This doesn't really suprise me. We've been using db2 in production on Linux at work (www.osogrande.com) for about 2 or 2 1/2 years (since the UDB 5 beta on Linux came out). It is easy to install (aside from some curses incompatibilities on RH7.1, which will get resolved shortly), easy to administer and performs well.
For free databases, I prefer postgres (transactional support, referential integrity, triggers, etc.). For commercial support, I have trouble liking Oracle (terrible query optimization for large queries, no statistics-based optimization like db2, much harder to administer, etc.). DB2 UDB on Linux is a low-cost, high-feature, high-performing dream.
Standard example: WalMart logs every single purchase ever made in order to make data warehouse analyzations, like finding what items are often bought together so that they can also be put on display next to each other to get a few more people to buy them together. That's several million entries each day...
The illegal we do immediately. The unconstitutional takes a little longer.
--Henry Kissinger
I want M$ to repeat their test with four 4-cpu machines running for a fair comparison. Whine, whinge, whimper.
... something a lot of
the posters here seem to have missed. To be of any real use, the TPC results have to be grok'd in fullness ... not just taken at headline value.
While the benchmarks aren't meant for direct comparison, we all use them that way. If those using them in such comparisons cannot make the needed "fudge factor" adjustments to do so, then they shouldn't be making them in the first place.
The benchmarks are not intended to relieve their users of the need to think
Might as well shoot your self now, that will never happen. Just to be on the safe side, take a jacket incase hell gets cold.
NCR. Founded 1885. Headquarters in Dayton, Ohio. Over a century in the "point of sale" business. And now #1 in transaction processing. Now that's an old-economy company.
Regards, Tommy
The test was run on SGI hardware. Even though it's intel, it's still SGI. I love the SGI hardware I've been exposed to, but you have to go into it with the knowledge and acceptance of the plain fact that you will pay out the nose for it and maybe a few other orfii too if the sales weasel is being efficient that day. At least they have the decency to make their uberexpensive machines look nice.
So while the OS was free, the hardware was cost++. (think about it man, 16 p!!! zeons alone ...)
--
News for geeks in Austin: www.geekaustin.org
News for Geeks in Austin, TX
How many companies are there out there that really have 3TB databases? Not too many I would guess. What I would be more interested in is how robust Linux would be in running multiple small databases (between 10 to 50 Mb). Kind of like "mass virtual hosting" except as a database server. I think this sort of benchmark would be much more valuable and give companies like ours just one more reason to throw away the burden of Microsoft products...
Nathaniel P. Wilkerson
Domain Names for $13
Nathaniel P. Wilkerson
www.haidacarver.com
TPC has more than 1 chart per benchmark, mostly 1 for raw performance, and 1 for price/performance. Check for example www.tpc.org and click on 'benchmarks' at the left. Select a benchmark from the list and you can select from the charts which one you prefer. In the end, it's the reader who decides WHAT's important, not the benchmark itself.
--
Never underestimate the relief of true separation of Religion and State.
It's interesting to be in the computer field lately. First Juniper starts challenging Cisco's market and profits, the DoD convicts microsoft of being a monopoly, then most dot-coms die (Eazel included) and now, it looks like Oracle's stronghold is being tred on by IBM's DB2 (quite an interesting war if you read all the info on it) and MySQL (and other open databases). I just don't know what monopolies still around to be crushed next.
Slashdot gets worse every day... Pipedot: News for nerds, without the corporate slant
You're not your fucking kahki's
second society
Hey, who needs those damn Steroids? Get a 4 ounce bottle of cough syrup, and it will be very hard for anyone to hurt you. Of course, they can always win the fight by pointing out how that one cloud looks just like a dolphin, and let you mull over that for the next 4 hours.
Hopefully I didn't put any [] around my words.
For some reason, this reminds me of what happens when a martial artist trains for 5 or 6 years, going through over a thousand practice fights and then gets jumped on the street. The first time they hit someone and actually see that person go down, it is always a little bit of a surprise.
Even though everyone who has used Linux , or worked on Linux, knows how great it is, when we all realize that our little pet OS can really knock out the people way bigger then us, we are always a little surprised...
Of course, since my experiences both with getting jumped and trying to use Linux are both slightly unsuccesful, maybe I shouldn't be the one to comment... ;)
Hopefully I didn't put any [] around my words.
Good point -- to be truly fair, the "correct" approach to adding more hardware would be to construct a cluster which would end up costing much more.
But my point was that for DSS, you probably don't need a cluster. You probably can get by with data replication to your Nth database server. If your sole purpose is to support a large number of ad-hoc DSS queries on historical data, your server don't even need to be networked to each other (as long as you have some way of updating the data on all servers).
It's not a fair comparison; but for DSS, the SGI/Linux score isn't as impressive at it could have been.
BTW, the Sequent (IBM) NUMA-Q that score well in the higher classes are basically 4-processor servers clustered together with optical internetworking. They are essentially a cluster sold as a "partitionable" single computer!
While I'm happy to see Linux on the TPC-H leader board, there are several things that makes me feel this "win" isn't as glorious as it could be...
First of all, the cost/performance ratio is much higher than the Windows-based entrants. Since this is a decision support benchmark, the work is heavily skewed to reading mostly static data; a solution which more easily scales up by throwing hardware at it.
Put differently, if you run two clusters of the 2nd (3rd, 4th or 5th) ranked system, and it'll have better performance than the Linux solution, but at lower cost/performance ratio.
Second, the results show that Windows (coupled with cheap hardware) is working so well at the low end that it pretty much displaced Unix systems. (Geez, no wonder Sun's getting creamed in the desktop workstation market these days.)
Third, this "win" is not for a write-heavy database application -- in my personal experience, Linux has been an underperformer in write-heavy database application. (With synchronous writing turned on to guarantee data flush to disk). This makes it less appealing for real-time transactional database work.
Separately, I wish such tools as PowerBuilder and CrystalReports were available for Linux... That would go a longer way to making Linux acceptable to the Enterprise...
Take a pill!
I love the sound of that. It brings up images of things like "well, we just spent half a million dollars on this new hardware and database software, I sure hope that decision is justified"
--
Friends don't let friends use multiple inheritance.
We own piles of IBM servers, some of which come with DB2 licences, so we thought we might as well use them. Plus, if the load gets too much for our little 4 CPU Netfinity box, we can move up to an AS400 without worrying about porting issues.
I'd be much more enthusiatic about PostgreSQL if it was a 'proper' OODB - methods on classes, collections rather than tables and all that. It seems to be a solid database, now they need to add the nifty features that will make it stand out from the crowd.
Don't discount the value of a decent stored-procedure language - there's a serious impedance mismatch between the low level systems languages (certainly C, and to an extent Java) and more SQL-like declarative languages.
Two of my main uses for stored procedures are parameter-driven queries (which have some complex processing to decide what query to run) and data munging. Both of these cases are query/update bound, so you won't benefit much from the improved performance of a compiled language, and you have to spend all that extra time dealing with the baggage, creating variables, #including, sizing variables, pre-processing, compiling, installing, etc. Not to mention if (like I used to do) you have to support around a dozen platforms, each with their own cc wrinkles. And then there's VAXen...
In this situation a procedural SQL language helps quite a bit, and Oracle's PL/SQL is a gazillion times better than T/SQL (SQL Server) or DB2's procedural SQL solution.
PL/SQL has proper exception handling, DB2 has a sort of ON ERROR GOTO, not as good as proper exception handling but does the job. T/SQL's error handling is right up there with C - an errorlevel variable, which means you have an ugly mix of error handling and normal code.
In PL/SQL you can declare variables to be of table.column%type, so if the table definition changes (say you make a varchar column longer), stored procedures will automagically keep on working. Saves some MAJOR maintenance headaches if you use this properly. SQL Server lets you declare your own data types, but you would end up with hundreds of the things, practically one for every column in every table, which is complete overkill. DB2 doesn't seem to have the concept of column data types.
The source for a PL/SQL stored proc can be stored in the database, so you don't have to worry about losing it. SQL Server does this, too (although I'm not sure if you can remove it for a production environment like Oracle can). DB2 pre-compiles the stored proc and doesn't keep the source (and, as I noted previously, sets the query plan at compilation time.), so you can't browse around a DB looking at the source code for the stored procs to figure out which version of your code they're running.
Oracle and DB2 have cursor loops, with automatically declared local vars for each column in a query. In SQL Server you're expected to DECLARE vars for each column, then FETCH columnlist INTO varlist. Imagine your query returns 50 columns - that's a whole pile of unnecessary code to maintain.
Also (and this REALLY bugs me) in DB2 when a stored proc returns a recordset, it DOESN'T RETURN THE COLUMN NAMES! I couldn't believe it to begin with, this means if you add or re-order the columns in a table, all your code (which has to use integers to index the columns) will be wrecked.
All this is IME, I may be wrong about some things
When the logs fill up (which incidentally happened a few minutes ago) we get a 'logs are full' message; (why can't we have dynamically re-sizing log files?) This is different - we force a 'run-away' job (no idea why it ran away), and the database stops.
Integrity is great, but the design of DB2 is such that it's too easy to get into a state of questionable integrity, at which point DB2 stops. Just run a LOAD statement and leave out the word 'UNRECOVERABLE' - bam! Tablespace is in backup pending state and you can't access the data. Hence the practice of 1 tablespace per table.
On-line backups have only just been added to the Unix version of DB2, so I expect they operate in quite a different way from OS/390; the documentation is very vague.
I'm not (nor do I want to be) a DBA, but I'll be doing some training when the IBM circus comes to town.
Thanks for the info on lock escalation - I'll look into that and see if that helps.
Handy, thanks.
hmm... everyone on comp.databases just trashes Mimer. If you insist on going closed-source, use Solid. Or wise up and use Postgres.
Some things I agree with and some things I don't
If you compare prices on a O3000 compared to a Starfire there prices are pretty clueful
Service/support, we have great support from SGI (of course we have lots and lots of servers). From their database side, what I had heard is that someone at SGI pissed off Larry Ellison at Oracle, so he's let SGI whither on the vine.
Poor commitment, yes when the bastard (who I shall not say) went to Microsoft was running things. I almost think he put the 320 & 520 into SGI's line just to get a job at Microsoft.
Which technology... for the 3000 we've got here I can get SCSI, FC, PCIx, Myranet, Gig E, OC3, FDDI, USB, Firewire, plus all of the special digital video stuff you want.
Longevity... no question here
Marketing has been a thorn in SGI's side ever since I've used them. Their Origin line was a kick ass server at the time, but everybody only thought of them as makers of movie effects (and still do to an extent)
The problem becomes from a price/performance metric as to what database do you use? The DB2 software itself was more expensive then the servers or the storage. 1/3 of the price of the entire system (300k+) just for the database software tends to eat into the price quickly, but I don't know of an alternative. When we checked out Oracle for Linux it was almost $200k for a single Compaq 2x PIII box, about the only option they had were MySQL or Postgress to reduce costs which would not have been up to the task.
Normally you don't want to make your filesystems that big anyway, from a backup/restore issue it becomes a killer. Any DBA worth his salt uses raw filesystems if possible, meaning you can only have one backup process on that partition i.e. normally meaning only one tape drive spinning for the entire partition. Retrieving from a single tape we can get 10-15 gig/hour, now for a restore of a 2 terabyte partition you'd be waiting a LONG while, before you could even think of applying your redo logs.
They have given an amazing level of commitment to IA32, but they have given even more to IA64 which will be very interesting IF they can pull it off. IA64 Linux numa clusters scalling to 1024 proc as a single image... that's what I call sweet.
I can deffinetly say that I like what SGI has done with the O3k series, it's cheaper than the O2K, it's a bit faster with the same mem & cpu (the faster interconnects give the boost), and I can just add an additional "p-brick" if I need more processing power without having to purchase all of the chassis stuff (XIO, disk spaces, etc). Not to mention the system partitioning capability, which I can't believe SGI slip through there fingers by allowing Sun to purchase the tech from Cray when they bought them... but that was when the Microsoft bastard was at the helm.
Glad I pay for my DSL service!
Dog is my co-pilot.
What is this? Programming or religious vendetta? They are a company you don't like -- so what? It's only a computer operating system. People are not dying because they get Windows on their machines, any more than they get a particular engine with their new Toyota.
Grow up.
- I don't care if they globalize against free speech. All my best free thoughts are done in my head.
Comment removed based on user account deletion
Comment removed based on user account deletion
Comment removed based on user account deletion
As predicted by the late Douglas Adams in his book "Dirk Gently's Holistic Detective Agency" - Slashdot is one of the first Electric Monks.
For those that don't know - an Electric Monk is a device which makes your life easier by believing things for you. To be any use they have to be pretty adept at believing several contradictory things at once.
I think a better reason might be that they're jealous of MS's success? Face it, if enough of their corporate customers demanded Linux software tomorrow, Gates and crew would have no compunctions in creating Office et al for Linux. Or even their own distro -- though they'd probably try to lock people into *that*... leopards rarely change their spots.
And, on a sidenote, don't think Sun, Oracle, et al are angels. (Though, the one company which probably best groks the culture of open source today is IBM.) They support Linux because it is convenient for them to do so. And Sun's support to Linux *on the SPARC platform* comes dangerously close to being called 'lip service'.
____________________________
2*b || !(2*b) is a tautology
My cubical resides next to a Microsoft consultant's. Everytime he sees one of my corporation's employees using a Microsoft product he says "Ya!". Like it is a win for him and M$.
For example he saw my coworking using IE - "ya!" he says. He sees another coworking using Microsoft MAP. "Ya!" he says, like he and Microsoft are winning an invisible game.
It is this aspect of Microsoft culture that I hate. Its like playing chess in highschool with someone who rubs it in after every move that he win. He cheats and then ridiculules you when he wins the game.
This permeates Microsoft culture and taints everything that they touch. They lie, they cheat, they bully and yes they win. It is not the technology that annoys me -which is mediocre, but good in terms of ease of use. It is Microsoft's competitive culture that has allowed it to become the neighbourhood bully.
The http server in the kernel blows away IIS in benchmarks. What about a database embedded in the kernel? for a dedicated db server the size of the kernel wouldn't matter too much. Is this feasable?
I don't hate MS, rather, I like the idea of a 'Free' product providing more value than the best commercial 'non-free' version of the same product. THat's what I'm happy about.
Keeping
First off, if your hardware doesn't suck, you shouldn't have this issue. But, I had it anyway with non-sucky hardware (256MB/Ath750) using the Mandrake 7.1 stock kernel. But a recompile fixed this issue. I suspect the problem was IDE related, but I'm not certain. Also, if you use XFree86 3.3.6, try upgrading to 4.0.x, it's faster for most (nVidia, ATi, 3dfx, Matrox, Trident, NeoMagic, Silicon Motion, 3DLabs) chipsets.
If violence isn't solving your problems, you're not using enough of it. - MAJ Misato Katsuragi
I want to congratulate the Linux community on the TPC win. As an avid fan of SQL Server 2000, I must respect the TPC results. However, I find it interesting that the TPC was said to be a worthless benchmark by most "M$" bashers until Linux came out on top.
Oh well, at least we can agree on something now.
"Saddam Hussein cavorts with terrorists."
No way! I was hating him long before the first Linux kernel ever shipped...
-----
free the mallocs!
I was quite surprised to find out many big companies are using linux. I think it just doesn't get a lot of publicity. I know of at least one Shell (the oil company) division that is using linux.
I think it makes sense business-wise to use linux in companies, it's a decent OS and it's free (Instead of paying for thousands of licenses).... The only problem is that secretaries and marketing people probably can't use it too well.
This makes me tempted set up DB2, then take up Oracle on it's million dollar challenge
/*drunk.. fix later*/
Come on. Moderate to strong agreement or strong disagreement with an article is practically the only criterion Slashdot "editors" use to decide what to post.
The only certainty is entropy.
DB2 is an old mainframe database pawned off as a "Modern Relational DB" which it is not. That's 90% of the reason why their stored proc handle errors like an old COBOL program.
Javascript + Nintendo DSi = DSiCade
Unisys modifed the NT code? arf arf arf!
Explain to me just what is so unbelievable about that? The windows source code CAN be licensed in case you weren't aware. They only do it for selected partners who are in bed with them and can pony up the cash. How the hell do you think Citrix created WinFrame in the first place!? THEY LICENSED THE CODE.
On top of that, it is painfully obvious that you have never even STUDIED OS programming. Why don't you go read up on my pack of lies, hm?
If I hadn't already posted I'd mod you down as a troll.
Javascript + Nintendo DSi = DSiCade
It sucks. Seriously, it sucks. Intel processors are very good at a single thread of execution (such as video games). As they are scaled up to multiple tasks you invariably have some form of hardware backlash due to the failure of decent on chip task switching. This may take the form of slower speeds, deadlocks, or processor expections (known to the windows world as 0E/0D/etc. blue screens of death).
Now speaking as someone who was actually at a conference in a Unisys facility when they announced this thing, they have problems. They took a standard Unix machine configuration of four processors per board with up to eight boards plugged into a gigabit crossbar and a shared memory architechture. WINDOWS NEVER DID THIS BEFORE! Unisys had to customize the NT code to make the damn thing even run, then covered it over with some figures about how much more I/O the machine could produce than a Sun Starfire E10000. But it is an inherently unstable design (at least for x86). Sun itself had a difficult time making 64 processors work. In fact in the UltraSparc III, there is on chip clustering of processors!
To make a long story short, Unisys did it to please Microsoft and found out that it wasn't practical.
Javascript + Nintendo DSi = DSiCade
The reason many people in the open source community dislike Microsoft so much is that their products don't "play well with others". The easiest example is .doc files. Any word processor that can't completely translate a .doc file is going nowhere fast in the business community (and Microsoft changes this undocumented format every couple of years) and this is just the tip of the iceberg. There was CORBA so they came out with COM. There was openGL so they came out with DirectX. There was Java so they came out with j++ and now C#. The list goes on and on. In each of those cases, the underlying fact is that if you chose the MS solution, you become locked down to their platform. And to open source people, freedom and choice are highly valued and Microsoft is constantly trying to marginalize anything that is not them or at least tied to them.
I miss the Karma Whores.
1st place $347us/QphH => Linux 2nd place $161us/QphH => Windows 2000 I am a *nix fan, I love my BSD (OpenBSD and FreeBSD) but this does nothing to dispel yesterdays slasdhot article from microfost on how windows is cheaper Total Cost of Ownership.... Chris Lee lee@mediawaveonline.com
Hmm, at $207 a QphH, the most economical thing seems like a farm or cluster of HP NetServer LXr 8500's.
Also, everyone is thinking this is a win for Linux. If you look at how the data is displayed, you'll notice operating system and actual system are both dependant upon one another and cannot be seperated. Thus, it is worthless to compare Linux running on a different system with that of Windows running on a different system. In all fairness, we don't even know what Windows will do on that system.
As it is, these rankings are meant to show the best arbitrary configurations which just happen to be setup somewhere and just happen to be submitted. Again, as others have noted, you cannot compare apples to oranges, so Linux being on top may be of little significance only that someone happened to have a system with Linux running on it. In terms of the defined QphH, it seems that server hardware is the biggest swayer, not the OS.
"I'll just chip in a bit for RedHat: I actually have that installed on my university machine." - Linus, '95
SShhh... I only saw that weird, half-baked, half-rendered HTML page 3 times so far. Of course, I've only been logged in for about 5 minutes.
Dancin Santa
My point exactly. The perceived toy-aspect of Linux is being shed. Linux is moving closer to the mainstream as evidenced by benchmarks such as this. It isn't the fact that Linux set some world record (dubiously), it's that Linux is represented as a viable alternative to Windows and other Unix systems.
Dancin Santa
There are a lot of problems with coming out and saying that Linux is the hands-down winner of this benchmark. The first problem is that the Linux system has twice as many processors as the next system down. The second problem is that the system costs twice as much as the next runner up. For these reasons alone it is foolhardy to immediately claim that Linux is now the undisputed heavyweight champion of the database world.
However, the story is more than that. The most important thing to notice is not that Linux is at the top, or the number of processors is so high, or even that the cost is exorbitant. The important thing is the Linux is a contender at all. This is an OS that hasn't until recently gotten a lot of respect. That Linux can "keep up with the big boys", it shows that it is certainly capable of handling the computing needs of the corporate community.
Of course, if you live by the benchmark, you die by the benchmark. Linux apologists would do well to acknowledge that it only breaks the world record because the hardware is twice as powerful as the next OS's. However, they can still crow about the significant progress that Linux has made over its relatively short lifetime. What other OS has gone from 0 to 60 in such a short time?
Dancin Santa
but a 3TB category! I realize that several companies have tons of data, but to create a category for 3TB DBs? Holy crap!
Sorry, I must not be conversant in this area, but wow!
Vote monkeys into Congress. They are cheaper and more trustworthy.
That SGI Machine is actually 4 machines.
4CPUs per machine, that makes it 16 CPUs in a clusters.
And Win2K is *still* the price/performance record holder.
--
Two witches watched two watches.
Which witch watched which watch?
Interesting that the SGI/Linux/DB2 combo had half the load time of the highest W2K/SQL7 server.
What is pirate software? Software for inventory of stolen treasure?
Checking both 100GB and 300GB mark, you'll notice that non of the MSQL+WIn2k go beyond 1700 QphH mark.
I would only agree with the above viewpoint if either
1) Microsoft was open source, or
2) Microsoft was a benevolent dictator
Neither is the case. Microsoft is pushing a closed, proprietary system with the goals of making as much money as possible, even if it means drastically reducing peoples' quality of life and compromising their rights on a widespread scale. In my opinion most companies operate on the same bottom line basis as Microsoft, and would be attempting to perpetrate similar ills on the world if they had the means... but Microsoft is among the very few companies holding the reigns on a highly closed system in a monopoly on a product that does not have much elasticity of demand (from a realistic corporate or home use perspective). Microsoft is further assisted by the fact that most people don't even understand the upstream issues; they simply endure a lousy end-result experience.
- First they ignore you, then they laugh at you, then ???, then profit.
Good SQL developers get big bucks, So of course IBM is going to sell it.
"BSD is about people pissing each other.." (Moid Vallat)
SGI submits it's workstation to the TPC with Linux as OS of choice.
Or something like that.
Rome taught me patience and assiduous application to detail. Virtues which temper the boldness of great, general views.
For all the Linux hype IBM would prefer to sell AIX and mainframe updated and minicomputer service contracts. They will do everything they can to keep you on IBM hardware, only if you call their bluff and say that you will leave IBM will they take your request to move to linux seriously.
Hence no linux benchmarks but as someone else has commented they would rather invest in doing a dynix benchmark.
holding a monopoly position in desktop OS marketshare
So now Linux doesn't exist as a desktop OS? Surely the 1,000 posts every hour praising the penguin prove differently?
The Linux community takes pleasure in being 'the other guys', and play on this whenever possible: 'Look at us! We're using free software, and we're doing things!', but when someone mentions microsoft, you start saying how unfair that microsoft is much bigger and more powerful: 'we're so small! microsoft is so unfair!'.
Make your mind up.
Stop blowing everything linux does out of proportion. Face the facts. Someone made a stupidly powerful machine, put Linux on it, and funnily enough, it won. If you put Windows 2000 on that computer, it would piss on Linux.
It wasn't the OS, but the hardware.
I could make an OS out of bits of tape and post-its, put it on a computer like that and it could out-perform anything else - it's not hard.
It's just a form of Penis envy!
And Walmart would wish to sponsor this event.
If you want a database that's really easy to administer, performs with the best of them, follows the SQL-99 standard, runs on Linux and is affordable, take a look at mimer.com for a free download.
About time you Linux weenies put your money where you're mouth is and submitted a result. However, note that cluster has 16 700MHz Xeons, whereas the Compaq Proliant entry below it has only 8. Look at the cost per transaction. More than twice as much.... Gee I thought you guys kept saying Linux has a lower TCO... And also note that when you compare the Windows2000 results the scalability is almost exactly linear. Number of CPUs and their speeds is linearly proportional to performance. This is also reflected in the price per transaction. But don't you guys keep saying Windows2000 doesn't scale... Looks like you're full of something.
Yes, but that's the expensive SGI hardware, not the OS. Perhaps Compaq or Dell would care to sponsor a benchmark?
Does a Christian soccer team even need a goalkeeper?
There is one thing that worries me about these results. All of IBM's entries for the category under 100 GB where running Windoze and SQL Server. Now this is sort of strange for a company with their own OS (OS/2) and their own Database (DB/2). Unless they were trying to pass the message that they just do not rely enough on their own systems. IBM must have felt ridiculous when they where beaten by an OS they have commited too and by their own product! Maybe next benchmark we will see more diversity in this range.
My other OS is the MCP!
You're words can be used in any number of situations. It's only a strip search. People are not dying from it. It's only taxation without representation. People are not dying from it.
No, people are not dying from having an inferior product shoved down their throats by a lucky monopolist. People are not dying because they lost their jobs at companies who tried to compete with a monopolist who forces competitors from the market using extortionist practices. People are not dying because choice in the marketplace has been destroyed.
But that doesn't make any of it right, fair or just. And just because the monopolist can do these things legally doesn't mean that I can't gloat and maybe even push a little when I see them stumble.
Maybe it has something to do with the situation I found myself in when I first started programming. Being a new programmer, I would make a lot of mistakes, and even though it was just a DOS program, every time I would hit one of my bugs Windows 3.1 would become completely unstable and require a complete reboot. Under suggestion from a friend, I switched to OS/2 and found programming much easier. When I hit a bug, all I had to do was close the DOS box and continue working.
Unfortunately, Microsoft's goal of 'one world - one OS' made it very difficult to use MY OS of choice for other things, since there was not much support from ISVs and hardware vendors. I detested all the FUD and underhanded tactics used by MS to destroy all competing OS solutions. It burned my soul that people would settle for a second rate solution just because they knew of no other, which they could not learn of due to Microsoft's tying deals and manipulation of the network effects.
I want to see MS fall. I want to see them fall hard. I want to see people fired for buying Microsoft, for when that day arrives every OS will have to stand on its merits, HW vendors will write portable device drivers, ISVs will write portable programs, diversity will thrive and reliability will overrule mediocrity. Until then , every nitch in Microsoft's armor is a pleasure to me.