Here's the sequence of steps I've seen... 1. The salesman who bids RAID5 configuration wins the business, as per Dennis's story. He or she wins because the configuration requires fewer disks than the alternative RAID10 configuration. The salesman gets a nice commission and goes to his company's sales club.
2. Since the system was "sized" for storage capacity (byte counting) instead of I/O rate capacity (I/O-per-second counting), the system runs the risk of failing to keep up with I/O throughput requirements. Especially because RAID5 configurations perform more I/O operations than you think for every Oracle block written by DBWR. 3. If the system has a high enough I/O rate, the company that bought the RAID5 configuration finds out the hard way that the I/O subsystem is severely undersized. The total price of the corrected configuration is more than if the company had bought the RAID10 configuration to begin with. It's a hard deal. BAARF. Cary Millsap Hotsos Enterprises, Ltd. http://www.hotsos.com Upcoming events: - Hotsos Clinic 101 in Dallas, Washington, Denver, Sydney - Hotsos Symposium 2004, March 7-10 Dallas - Visit www.hotsos.com for schedule details... -----Original Message----- Mladen Gogala Sent: Sunday, June 15, 2003 2:24 PM To: Multiple recipients of list ORACLE-L Dennis, to tell the truth, writing in oracle is not a big problem, as long as the redo files are not on RAID-5. Everything else can reside on RAID-5 without a visible performance impact. Second, RAID-5 vendors like EMC and Hitachi usually offer two versions of non-volatile cache: write-through one which essentially performs prefetch and a genuine full cache which caches both read and write calls. The latter type of cache, which is very expensive, is found on Symmetrix boxes only and not on former DG-Clariion boxes (talking EMC here). These types of RAID-5 implementation are usually referred to as RAID-6 or RAID-S. How to benchmark those? Well, the trick in benchmarking those systems is to do what one would never do with it's own system: put redo logs on RAID-5(6,S?), launch several threads of update intensive short transactions (OLTP mix) and count "user commits" from v$sysstat. Prior to that, establish a baseline with RAID 1+0 and see what is the difference. See how many commits would RAID-5 box record during the same time as RAID-1+0 box and you'll know the difference in speed. Also, make sure to pull out one of the disks while system is working and see what's the impact of resilvering. <RANT> As for the entertainment value, I would hope that Julia Roberts and Mel Gibson would consider making a movie about the RAID-5 conspiracy. Julia would be a DBA trying to purchase a RAID box and Mel Gibson would be a honest RAID-5 salesman which would uncover a nasty EMC, IBM and Hitachi conspiracy. You can tell that it is a fiction because of the phrase "honest RAID salesman". The only problem would be to teach the two of them how not to sound "nucular". </RANT> On 2003.06.15 14:14, DENNIS WILLIAMS wrote: > Mladen - I would propose another theory. Please consider this rant for > entertainment value, and hopefully to consider a book on this topic. > Usually companies request bids from several vendors. If you as a storage > salesperson always bid RAID1+0, you will always be underbid by your > competitors, receive no commissions, and in the end be fired. If you always > bid RAID5, you will probably receive some business, some commissions, and > keep your job. Now, pretend I'm a storage system salesperson. Which system > am I going to specify to the customer? > Now, if the customer insists on purchasing RAID1+0 or whatever, I'll > probably argue a little because I've been stung before where I was suckered > into bidding RAID1+0 and then my ignoramus competitor just blindly quoted > RAID5 and when it bubbled up to the V.P. he asked "why are you going with > the more expensive vendor"? And trying to quote a salesperson isn't a good > move at that point. > I thought Rachel had a good point on this topic awhile back. If you are a > top-notch consultant that is often called in to solve performance problems, > you have often cured them by switching from RAID5. So you have a lot of > confidence in it. But if you are the lowly on-site DBA just trying to hang > onto your job in the political turbulence, you usually don't have enough > facts to challenge the system administrators who trust their storage vendors > a lot more than a DBA that couldn't actually configure a storage system to > save his or her life. > So suppose I do try to challenge the system administrators. I say that > RAID1+0 will write faster than RAID5. The reply is "okay how much faster?" I > mumble that I've never seen any actual figures published anywhere. He or she > says maybe at the low level there is a slight advantage to RAID1+0, but with > a gig of battery-backed cache that won't be true. Or he or she asks "What > does Oracle recommend?". And on it goes. > Okay, I'm being provocative here. But how do I, a lowly DBA, prove which > is faster? Should I talk my system administrator into going to the trouble > of configuring a system both ways and run some tests? What type of tests > would be most useful? If I choose the wrong test, and RAID5 looks just as > good as RAID1+0, I'm sunk. Then for years to come when I try to make a point > at a meeting someone will say "yeah, is this another RAID1+0 theory?" > > Dennis Williams > DBA, 80%OCP, 100% DBA > Lifetouch, Inc. > [EMAIL PROTECTED] > > > -----Original Message----- > Sent: Sunday, June 15, 2003 2:49 AM > To: Multiple recipients of list ORACLE-L > > > RAID-5 microkernel has more lines of code then Oracle7? Wow! This is an > astonishing piece of information and if it wasn't coming from you, I'd > dismiss it as yet another conspiracy theory. > > On 2003.06.15 01:44, Cary Millsap wrote: > > > Meanwhile I have never understood why storage vendors would prefer > > > selling RAID5 over RAID10. > > > > ...Because if they don't sell RAID5, they don't recover the R&D costs of > > creating a RAID5 microkernel that has more lines of code than the Oracle7 > > executable. I'm not kidding. > > > > > > Cary Millsap > > Hotsos Enterprises, Ltd. > > http://www.hotsos.com > > > > Upcoming events: > > - Hotsos Clinic 101 in Dallas, Washington, Denver, Sydney > > - Hotsos Symposium 2004, March 7-10 Dallas > > - Visit www.hotsos.com for schedule details... > > > > > > -----Original Message----- > > Niall Litchfield > > Sent: Thursday, June 12, 2003 12:05 PM > > To: Multiple recipients of list ORACLE-L > > > > Jared writes > > > At the meeting last week I wore my 'No RAID 5' hat. > > > > > > Those of you at IOUG 99 in Denver may have seen it, I > > > wore it every day there. ;) > > > > I'm curious now. Pictures required. > > > > Meanwhile I have never understood why storage vendors would prefer > > selling RAID5 over RAID10. More disks=more profit surely? Also 10 > 5 > > therefore self evidently twice as good for all applications. Meanwhile > > we have this strange situation where performance consultants are > > publicising the fact that you have less need for performance consultants > > with RAID10 than with RAID5. > > > > Niall > > > > -- > > Please see the official ORACLE-L FAQ: http://www.orafaq.net > > -- > > Author: Niall Litchfield > > INET: [EMAIL PROTECTED] > > > > Fat City Network Services -- 858-538-5051 http://www.fatcity.com > > San Diego, California -- Mailing list and web hosting services > > --------------------------------------------------------------------- > > To REMOVE yourself from this mailing list, send an E-Mail message > > to: [EMAIL PROTECTED] (note EXACT spelling of 'ListGuru') and in > > the message BODY, include a line containing: UNSUB ORACLE-L > > (or the name of mailing list you want to be removed from). You may > > also send the HELP command for other information (like subscribing). > > > > -- > > Please see the official ORACLE-L FAQ: http://www.orafaq.net > > -- > > Author: Cary Millsap > > INET: [EMAIL PROTECTED] > > > > Fat City Network Services -- 858-538-5051 http://www.fatcity.com > > San Diego, California -- Mailing list and web hosting services > > --------------------------------------------------------------------- > > To REMOVE yourself from this mailing list, send an E-Mail message > > to: [EMAIL PROTECTED] (note EXACT spelling of 'ListGuru') and in > > the message BODY, include a line containing: UNSUB ORACLE-L > > (or the name of mailing list you want to be removed from). You may > > also send the HELP command for other information (like subscribing). > > > > -- > Mladen Gogala > Oracle DBA > -- > Please see the official ORACLE-L FAQ: http://www.orafaq.net > -- > Author: Mladen Gogala > INET: [EMAIL PROTECTED] > > Fat City Network Services -- 858-538-5051 http://www.fatcity.com > San Diego, California -- Mailing list and web hosting services > --------------------------------------------------------------------- > To REMOVE yourself from this mailing list, send an E-Mail message > to: [EMAIL PROTECTED] (note EXACT spelling of 'ListGuru') and in > the message BODY, include a line containing: UNSUB ORACLE-L > (or the name of mailing list you want to be removed from). You may > also send the HELP command for other information (like subscribing). > -- > Please see the official ORACLE-L FAQ: http://www.orafaq.net > -- > Author: DENNIS WILLIAMS > INET: [EMAIL PROTECTED] > > Fat City Network Services -- 858-538-5051 http://www.fatcity.com > San Diego, California -- Mailing list and web hosting services > --------------------------------------------------------------------- > To REMOVE yourself from this mailing list, send an E-Mail message > to: [EMAIL PROTECTED] (note EXACT spelling of 'ListGuru') and in > the message BODY, include a line containing: UNSUB ORACLE-L > (or the name of mailing list you want to be removed from). You may > also send the HELP command for other information (like subscribing). > -- Mladen Gogala Oracle DBA -- Please see the official ORACLE-L FAQ: http://www.orafaq.net -- Author: Mladen Gogala INET: [EMAIL PROTECTED] Fat City Network Services -- 858-538-5051 http://www.fatcity.com San Diego, California -- Mailing list and web hosting services --------------------------------------------------------------------- To REMOVE yourself from this mailing list, send an E-Mail message to: [EMAIL PROTECTED] (note EXACT spelling of 'ListGuru') and in the message BODY, include a line containing: UNSUB ORACLE-L (or the name of mailing list you want to be removed from). You may also send the HELP command for other information (like subscribing). -- Please see the official ORACLE-L FAQ: http://www.orafaq.net -- Author: Cary Millsap INET: [EMAIL PROTECTED] Fat City Network Services -- 858-538-5051 http://www.fatcity.com San Diego, California -- Mailing list and web hosting services --------------------------------------------------------------------- To REMOVE yourself from this mailing list, send an E-Mail message to: [EMAIL PROTECTED] (note EXACT spelling of 'ListGuru') and in the message BODY, include a line containing: UNSUB ORACLE-L (or the name of mailing list you want to be removed from). You may also send the HELP command for other information (like subscribing).
