raid1 question
Hi All, I'm using a raid1 setup with the raidtools 0.90 and mingo's raid patch against the 2.2.15 kernel. My question is how does the raid driver decide which disk in the mirror to use as the source for synchronization when mkraid is used? I tried a few experiments to see how it behaved. The /etc/raidtab is as follows: raiddev /dev/md0 raid-level 1 nr-raid-disks 2 nr-spare-disks 0 chunk-size 4 persistent-superblock 1 device /dev/sdb1 raid-disk 0 device /dev/sdc1 raid-disk 1 My first experiment was to shutdown, unplug /dev/sdc1, and bring up the system again. The raid driver happily ran the mirror in degraded mode. I wrote some files onto /dev/md0, rebooted the machine, with /dev/sdc1 plugged back in. /dev/md0 continued to run in degraded mode until I ran mkraid to resynchronize the mirror. The files that were written to /dev/md0 while it was running in degraded mode were preserved. :) I then tried the same experiment, but this time unplugged /dev/sdb1 so /dev/sdc1 (which became /dev/sdb1) ran in degraded mode after the reboot. I wrote some files onto /dev/md0 while it was in this state. After plugging the original /dev/sdb1 in after a reboot, the mirror continued to run in degraded mode, but this time using /dev/sdc1, because the persistent superblock was more up to date. I then ran mkraid to resynchronise the mirror again, but this time, the files written to /dev/md0 in degraded mode (really /dev/sdc1) had disappeared! :( So, when /dev/md0 is set up with raid1, does it always use raid-disk 0 as the source for a resync? My concern is if /dev/sdb1 really crashes and I replace it with another fresh disk, partition it the same as before, and do a resync, everything on /dev/sdc1 (raid-disk 1) will be deleted. If this is the case, I presume the solution would be to change the SCSI id of /dev/sdc1 so it becomes /dev/sdb1 in the mirror, making it raid-disk 0? Thanks, Ben.
IDE Controllers
I was wondering which add in PCI IDE controllers are good to use and SMP safe with a 2.2.14 or 2.2.15 kernel. I did some looking for Ultra66 controllers and the only thing I could find that was supported was the Ultra66 from Promise. Checking on their site, they state the the driver is included with 2.2.10 and later, but in 2.2.15 I could not find a reference to an Ultra66 driver in the kernel sources (or I was being very stupid :-) ). Plus, the driver that I downloaded from Promises website states that it isn't SMP safe. So to sum up a long story. I have four drives, using RAID and would like to migrate the two slaves to their own controllers (making them single/masters of course), both for performance reasons as well as RAID stability reasons. What PCI IDE controllers have good support in the stable (2.2.14/2.2.15) tree? -- Edward Muller [EMAIL PROTECTED] [EMAIL PROTECTED]
RE: performance limitations of linux raid
(I really hate how Outlook makes you answer in FRONT of the message, what a dumb design...) Well, without spending the time I should thinking about my answer, I'll say there are many things which impact performance, most of which we've seen talked about here: 1 - how fast can you get data off the media? 2 - related - does the data just happen to be in drive cache? 3 - how fast can you get data from the drive to the controller? 4 - how fast can you get data from the controller into system RAM? 5 - how fast can you get that data to the user? (assuming reads - writes are similar, but reversed - for the most part) Number 1 relates to rotational delay, seek times, and other things I'm probably forgetting. (Like sector skewing (is that the right term? I forget!) - where you try to put the 'next' sector where its ready to be read by the head (on a sequential read) just after the system has gotten around to asking for that sector (boy, I can see the flames coming already! ;-) / 2 Number 2 relates to how smart your drive is, and too smart a drive can actually slow you down by being in the wrong place reading data you don't want when you go ask it for data somewhere else. Number 3 relates to not only the obvious issue of how fast the scsi bus is, but how congested it is. If you have 15 devices which can sustain a data rate (including rotational delays and cache hits) of 10 megabytes/sec, and your scsi bus can only pass 20 MB/sec, then you should not put more than 2 of those devices on that bus - thus requiring more and more controllers... (and I'm ignoring any issues of contention, as I'm not familiar with the low level of scsi enough to know about it) Number 4 relates to your system bus bandwidth, DMA speed, system bus loading, etc. Number 5 relates to how fast your cpu is, how well written the driver is, and other things I'm probably forgetting. (Like , can the OS actually HANDLE 2 things going on at once - and do floppy accesses take priority over later requests for hard disk accesses?) So maximizing performance is not a 1-variable exercise. And you don't always have the control you'd like over all the variables. And paying too much attention to only one while ignoring others can easily cause you to make really silly statements like: "Wow, I've really got a fast system here - I have an ULTRA DMA 66 drive on my P133 - really screams. And with that nice new 199x cdrom drive with it as secondary - wow, I really SCREAM through those CD's!" Um, well, sure, uh-huh. Most of you on this list see the obvious errors there - I've seen some pretty smart people do similar things (but not so obvious to most folks) by missing some of the above issues. Well, this is way longer than I expected, so I'll quit before I get into any MORE trouble than I probably am already! rusty -Original Message- From: Bob Gustafson [mailto:[EMAIL PROTECTED]] Sent: Thursday, May 04, 2000 5:18 PM To: [EMAIL PROTECTED]; [EMAIL PROTECTED] Subject: Re: performance limitations of linux raid I think the original answer was more to the point of Performance Limitation. The mechanical delays inherent in the disk rotation are much slower than the electronic or optical speeds in the connection between disk and computer. If you had a huge bank of semiconductor memory, or a huge cache or buffer which was really helping (i.e., you wanted the information that was already in the cache or buffer), then things get more complicated. BobG
Re: IDE Controllers
On 5 May, Edward Muller wrote: I was wondering which add in PCI IDE controllers are good to use and SMP safe with a 2.2.14 or 2.2.15 kernel. I did some looking for Ultra66 controllers and the only thing I could find that was supported was the Ultra66 from Promise. Checking on their site, they state the the driver is included with 2.2.10 and later, but in 2.2.15 I could not find a reference to an Ultra66 driver in the kernel sources (or I was being very stupid :-) ). Plus, the driver that I downloaded from Promises website states that it isn't SMP safe. I can't tell you about the SMP support, but the driver is NOT included in the standard source tree. You need to get your hands on the IDE patch from "The Linux IDE Guy". All his work is included in the 2.3.xx and should be in the 2.4 kernel (much like the RAID). This card has worked well for me, but I have yet to try it in a software RAID. Using the "new"RAID patches. What PCI IDE controllers have good support in the stable (2.2.14/2.2.15) tree? Check out the patch, there are at least 2 other cards. http://www.kernel.org/pub/linux/kernel/people/hedrick/ide.2.2.15.2504.patch.bz2 -- Matt Valites([EMAIL PROTECTED]) The Axium - http://www.axium.net The Internet is full. Go away.
RE: performance limitations of linux raid
From: Gregory Leblanc [mailto:[EMAIL PROTECTED]] ..., that would suck up a lot more host CPU processing power than the 3 SCSI channels that you'd need to get 12 drives and avoid bus saturation. not to mention the obvious bus slot loading problem ;-) rc
Re: performance limitations of linux raid
Not entirely, there is a fair bit more CPU overhead running an IDE bus than a proper SCSI one. A "fair" bit on a 500mhz+ processor is really negligible. Ehem, a fair bit on a 500Mhz CPU is ~ 30%. I have watched a *single* UDMA66 drive (with read ahead, multiblock io, 32bit mode, and dma transfers enabled) on a 2.2.14 + IDE + RAID patched take over 30% of the CPU during disk activity. The same system with a 4 x 28G RAID0 set running would be .1% idle during large copies. An exactly configured system with UltraWide SCSI instead of IDE sits ~ 95% idle during the same ops. Try turning on DMA
No Subject
Re: IDE Controllers
http://www.linux-ide.org/ Yes PDC20246/PDC20262 are SMP safe. On Fri, 5 May 2000, Edward Muller wrote: I was wondering which add in PCI IDE controllers are good to use and SMP safe with a 2.2.14 or 2.2.15 kernel. I did some looking for Ultra66 controllers and the only thing I could find that was supported was the Ultra66 from Promise. Checking on their site, they state the the driver is included with 2.2.10 and later, but in 2.2.15 I could not find a reference to an Ultra66 driver in the kernel sources (or I was being very stupid :-) ). Plus, the driver that I downloaded from Promises website states that it isn't SMP safe. So to sum up a long story. I have four drives, using RAID and would like to migrate the two slaves to their own controllers (making them single/masters of course), both for performance reasons as well as RAID stability reasons. What PCI IDE controllers have good support in the stable (2.2.14/2.2.15) tree? -- Edward Muller [EMAIL PROTECTED] [EMAIL PROTECTED] - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.tux.org/lkml/ Andre Hedrick The Linux ATA/IDE guy
Re: IDE Controllers
Yo Andre! 2.2.14 did not work for me. I have a dual PIII with onboard UDMA33 controller running RAID1. Very stable. When I just moved the two drives to a Promise Ultra66 the system became very unstable (uptimes in minutes). YMMV. RGDS GARY On Fri, 5 May 2000, Andre Hedrick wrote: http://www.linux-ide.org/ Yes PDC20246/PDC20262 are SMP safe. On Fri, 5 May 2000, Edward Muller wrote: I was wondering which add in PCI IDE controllers are good to use and SMP safe with a 2.2.14 or 2.2.15 kernel. I did some looking for Ultra66 controllers and the only thing I could find that was supported was the Ultra66 from Promise. --- Gary E. Miller Rellim 20340 Empire Ave, Suite E-3, Bend, OR 97701 [EMAIL PROTECTED] Tel:+1(541)382-8588 Fax: +1(541)382-8676
Re: IDE Controllers
RUDE surprize for you Hardware RAID 1 under promise is not hardware! Details.drives and host bios rev (promise)? On Fri, 5 May 2000, Gary E. Miller wrote: Yo Andre! 2.2.14 did not work for me. I have a dual PIII with onboard UDMA33 controller running RAID1. Very stable. When I just moved the two drives to a Promise Ultra66 the system became very unstable (uptimes in minutes). YMMV. RGDS GARY Andre Hedrick The Linux ATA/IDE guy
Re: IDE Controllers
Yo Andre! I know that, I am using Mingo's RAID-1 patch. Sort of figured that should be obvious on the linux-raid group... RGDS GARY On Fri, 5 May 2000, Andre Hedrick wrote: RUDE surprize for you Hardware RAID 1 under promise is not hardware! Details.drives and host bios rev (promise)? On Fri, 5 May 2000, Gary E. Miller wrote: Yo Andre! 2.2.14 did not work for me. I have a dual PIII with onboard UDMA33 controller running RAID1. Very stable. When I just moved the two drives to a Promise Ultra66 the system became very unstable (uptimes in minutes). YMMV. --- Gary E. Miller Rellim 20340 Empire Ave, Suite E-3, Bend, OR 97701 [EMAIL PROTECTED] Tel:+1(541)382-8588 Fax: +1(541)382-8676 - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.tux.org/lkml/
Re: How to remove a disk from Raidset which has not yet failed?
I have RedHat6.1, but raidsetfaulty doesn't seem to work for me. Am I missing something? My /sbin/raidsetfaulty is linked to /sbin/raidstart. /home/mcajalin/ftp_hdc/raid1]# cat /proc/mdstat Personalities : [raid1] read_ahead 1024 sectors md1 : active raid1 hdc8[1] hda8[0] 513984 blocks [2/2] [UU] md0 : active raid1 hdc7[1] hda7[0] 513984 blocks [2/2] [UU] unused devices: none /home/mcajalin/ftp_hdc/raid1]# ls -l /sbin/raidsetfaulty lrwxr--r-- 1 root root 15 May 5 10:36 /sbin/raidsetfaulty - /sbin/raidstart /home/mcajalin/ftp_hdc/raid1]# /sbin/raidsetfaulty /dev/md0 /dev/hda7 Unknown command /sbin/raidsetfaulty usage: raidsetfaulty [--all] [--configfile] [--help] [--version] [-achv] /dev/m d?* /home/mcajalin/ftp_hdc/raid1]# --- Neil Brown [EMAIL PROTECTED] wrote: raidsetfaulty /dev/mdx /dev/sdy1 NeilBrown __ Do You Yahoo!? Send instant messages get email alerts with Yahoo! Messenger. http://im.yahoo.com/
trouble with lilo on /dev/hdc
Hi there: My raid1 is running fine on /dev/hda(boot disk) and /dev/hdc. ;;Modified /etc/lilo.conf(Add one section for initrd.raid.img) Two cmds issued: mkinitrd /boot/initrd.raid1.img --with raid1 2.2.12-20 lilo -v ;;No warning. Then, power down, remove /dev/hdc, power up. /dev/hda is able to boot up. /proc/mdstat indicates it's running on partitions from /dev/hda only. Everything works as expected. (Partitions on /dev/hda and /dev/hdc are symmetrical) But if raid1 is running and /dev/hdc is boot disk, with same steps repeated I can't boot up. I see "01 01 01 ..." continuously running on screen. Same steps means the following: Modified /etc/lilo.conf from previous operation(from /boot/lilo.conf in /dev/hda, except "hda" was replaced by "hdc".) Two cmds issued: mkinitrd /boot/initrd.raid1.img --with raid1 2.2.12-20 lilo -v ;; get warning saying /dev/hdc is not the first disk. Then, power down, remove /dev/hda, power up. Can't boot up!! What am I missing here? Is "lilo -v" the problelm? Any help is appreciated. J. - Info with /dev/hda being boot disk. /etc]$ cat /proc/mdstat Personalities : [raid1] read_ahead 1024 sectors md0 : active raid1 hda7[0] 513984 blocks [2/1] [U_] md1 : active raid1 hda8[0] 513984 blocks [2/1] [U_] unused devices: none /etc]$ cat /etc/lilo.conf boot=/dev/hda map=/boot/map install=/boot/boot.b prompt timeout=50 default=linux_raid1 image=/boot/vmlinuz-2.2.12-20 label=linux initrd=/boot/initrd-2.2.12-20.img read-only root=/dev/hda11 image=/boot/vmlinuz-2.2.12-20 label=linux_raid1 initrd=/boot/initrd.raid1.img read-only root=/dev/hda11 cat /etc/raidtab # Config file for raid-1 device. raiddev/dev/md0 raid-level 1 nr-raid-disks 2 nr-spare-disks 0 chunk-size 4 persistent-superblock 1 device /dev/hdc7 raid-disk 0 device /dev/hda7 raid-disk 1 raiddev/dev/md1 raid-level 1 nr-raid-disks 2 nr-spare-disks 0 chunk-size 4 persistent-superblock 1 device /dev/hdc8 raid-disk 0 device /dev/hda8 raid-disk 1 - - Info with /dev/hdc as boot disk cat /etc/lilo.conf boot=/dev/hdc map=/boot/map install=/boot/boot.b prompt timeout=50 default=linux_raid1 image=/boot/vmlinuz-2.2.12-20 label=linux initrd=/boot/initrd-2.2.12-20.img read-only root=/dev/hdc11 image=/boot/vmlinuz-2.2.12-20 label=linux_raid1 initrd=/boot/initrd.raid1.img read-only root=/dev/hdc11 - __ Do You Yahoo!? Send instant messages get email alerts with Yahoo! Messenger. http://im.yahoo.com/
Re: IDE Controllers
What you do not know is that there will be a drive int the futre that will have a native SCSI overlay and front end. This will have a SCB-ATA converter/emulation. This will require setup and booting as a SCSI device. FUN, heh?? Andre Hedrick The Linux ATA/IDE guy
RE: IDE Controllers
-Original Message- From: Andre Hedrick [mailto:[EMAIL PROTECTED]] Sent: Friday, May 05, 2000 7:59 PM To: Gary E. Miller Cc: Linux Kernel; Linux RAID Subject: Re: IDE Controllers What you do not know is that there will be a drive int the futre that will have a native SCSI overlay and front end. This will have a SCB-ATA converter/emulation. This will require setup and booting as a SCSI device. FUN, heh?? Bleah, why? I haven't figured why there are all those IDE-SCSI hiding things yet. The history of Linux seems to point towards the IDE support being better than the SCSI, and yet the CD-R/W devices work through the SCSI interface, and it looks like now the disks will. Obviousally, I don't keep up on all of the kernel developments, I've still got a full time job to keep track of, but I'm still interested. Greg - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.tux.org/lkml/
Re: performance limitations of linux raid
On Fri, 5 May 2000, Michael Robinton wrote: Not entirely, there is a fair bit more CPU overhead running an IDE bus than a proper SCSI one. A "fair" bit on a 500mhz+ processor is really negligible. Ehem, a fair bit on a 500Mhz CPU is ~ 30%. I have watched a *single* UDMA66 drive (with read ahead, multiblock io, 32bit mode, and dma transfers enabled) on a 2.2.14 + IDE + RAID patched take over 30% of the CPU during disk activity. The same system with a 4 x 28G RAID0 set running would be .1% idle during large copies. An exactly configured system with UltraWide SCSI instead of IDE sits ~ 95% idle during the same ops. Try turning on DMA Ahem, try re-reading the above, line 3 first word! --- As folks might have suspected, not much survives except roaches, and they don't carry large enough packets fast enough... --About the Internet and nuclear war.