RE: RAID 5 versus mirroring

Eric, Yes, you can get the highest read speed out of RAID5. So it seems like for read databases it's the best one, but... You might get into a situation that happened 1 month ago with one of our Informix running severs (on RAID5). One of the disks started failing, no one noticed for couple days, then another one did as well, hot spare did not kick in time - so we got fried!!! There is no way to recover on RAID5 with 2 bad disks. And of course it took a while to restore from a backup, which is quite costly if you are running a production server and have no speedy failover strategy. From this point I am strongly taking the Art Kagel's side and saying : no more RAID5. ...just my 5c... Elena. -----Original Message----- From: Eric Fontaine [mailto:Efontaine@houston.rr.com] Sent: Friday, November 17, 2000 6:12 AM To: informix-list@iiug.org Subject: RAID 5 versus mirroring Ok guys, I've seen several posts regarding RAID 5 versus mirrored arrays. Maybe a disk guru can help me out here...I'm merely a lowly programmer. When running vmstat on our production dbserver, an RS6000, 6GB mem, 12 CPU's @ 262 Mhz with a software-mirrored array, 16 disks x 2, plus 2 for the OS, we're seeing the I/O wait part of the vmstat output in the 60-90% range. When we take a Level 0 backup and apply to another RS6000 with 2GB mem, and 4 340 Mhz CPU's, and 10 disks in a RAID 5 arrangement, we see vmstat showing 0-1% I/O wait. All machines are running AIX 4.3.1 with Informix 7.3.1 (which I think desperately needs to be upgraded, right?) So the question is - what exactly does the IO Wait part of vmstat mean, and is it indicative of poor disk I/O? Also intersting to note - running a single 4gl program on either box yields very similar performance results. Our db activity is mostly reads versus writes judging by looking at sysmaster information - bufreads, bufwrites, isam reads,writes, deletes - etc. When I sum everything up, reads outweigh writes by about 1000 to 1 or more on every table in the system. Hopefully this is a reasonalby accurate way to figure this? But I understand the read/write ratio is I've taken a look at how the DBA laid out tables on the mirrored machine, and only the largest tables (3 millions rows +) have fragments in multiple disks. For all the rest, the tables reside entirely on one physical disk. I suspect that a better layout would help out, but I don't know how much. Our management is asking for recommendations on ways to get more speed out of our machines. We can do whatever we want - hardware, software - whatever. But we need to spend money on the right stuff, of course. Any thoughts you can provide would be most appreciated - I've read lots of good info out here. Thanks!

Any thoughts from the group on the vmstat output? I was reading today and found something that said if you use raw devices (which we do), that disk I/O wait for raw devices isn't included in the vmstat I/O wait. So now I'm left to wonder - is this some kind of OS wait? And how do I find disk I/O wait for systems using raw devices? I'd like to know if we're having problems getting the data off the disks fast enough, or of the bottleneck is elsewhere. Elena makes a very good point about fail-safe of RAID 5....there's anyother good point to be made...when a RAID 5 disk fails, the controller or software must interpolate the missing or unavailable data by reading the remaining drives. If you have 30 disks in your array, the read overhead from a failed drive is huge, possibly bringing your system to it's knees, since you now have to read 29 other drives to interpret the missing data on the bad drive. "Elena Korol" <ekorol@styleclick.com> wrote in message news:8v43ke$732$1@news.xmission.com... > > Eric, > > Yes, you can get the highest read speed out of RAID5. So it seems like for > read databases it's the best one, but... > > You might get into a situation that happened 1 month ago with one of our > Informix running severs (on RAID5). > One of the disks started failing, no one noticed for couple days, then > another one did as well, hot spare did not kick in time - so we got fried!!! > There is no way to recover on RAID5 with 2 bad disks. > And of course it took a while to restore from a backup, which is quite > costly if you are running a production server and have no speedy failover > strategy. > > From this point I am strongly taking the Art Kagel's side and saying : no > more RAID5. > > ...just my 5c... > > Elena. > > > > -----Original Message----- > From: Eric Fontaine [mailto:Efontaine@houston.rr.com] > Sent: Friday, November 17, 2000 6:12 AM > To: informix-list@iiug.org > Subject: RAID 5 versus mirroring > > > Ok guys, I've seen several posts regarding RAID 5 versus mirrored arrays. > Maybe a disk guru can help me out here...I'm merely a lowly programmer. > > When running vmstat on our production dbserver, an RS6000, 6GB mem, 12 CPU's > @ 262 Mhz with a software-mirrored array, 16 disks x 2, plus 2 for the OS, > we're seeing the I/O wait part of the vmstat output in the 60-90% range. > > When we take a Level 0 backup and apply to another RS6000 with 2GB mem, and > 4 340 Mhz CPU's, and 10 disks in a RAID 5 arrangement, we see vmstat showing > 0-1% I/O wait. > > All machines are running AIX 4.3.1 with Informix 7.3.1 (which I think > desperately needs to be upgraded, right?) > > So the question is - what exactly does the IO Wait part of vmstat mean, and > is it indicative of poor disk I/O? Also intersting to note - running a > single 4gl program on either box yields very similar performance results. > > Our db activity is mostly reads versus writes judging by looking at > sysmaster information - bufreads, bufwrites, isam reads,writes, deletes - > etc. When I sum everything up, reads outweigh writes by about 1000 to 1 or > more on every table in the system. Hopefully this is a reasonalby accurate > way to figure this? But I understand the read/write ratio is > > I've taken a look at how the DBA laid out tables on the mirrored machine, > and only the largest tables (3 millions rows +) have fragments in multiple > disks. For all the rest, the tables reside entirely on one physical disk. > > I suspect that a better layout would help out, but I don't know how much. > Our management is asking for recommendations on ways to get more speed out > of our machines. We can do whatever we want - hardware, software - > whatever. But we need to spend money on the right stuff, of course. > > Any thoughts you can provide would be most appreciated - I've read lots of > good info out here. > > Thanks! >