Re: Software RAID (non-preempt) server blocking question. (2.6.20.4)
On Fri, 30 Mar 2007, Neil Brown wrote: On Thursday March 29, [EMAIL PROTECTED] wrote: Did you look at cat /proc/mdstat ?? What sort of speed was the check running at? Around 44MB/s. I do use the following optimization, perhaps a bad idea if I want other processes to 'stay alive'? echo Setting minimum resync speed to 200MB/s... echo This improves the resync speed from 2.1MB/s to 44MB/s echo 20 /sys/block/md0/md/sync_speed_min echo 20 /sys/block/md1/md/sync_speed_min echo 20 /sys/block/md2/md/sync_speed_min echo 20 /sys/block/md3/md/sync_speed_min echo 20 /sys/block/md4/md/sync_speed_min Yes, well You told it to use up to 200MB/s and the drives are only delivering 44MB/s, so they will be taking nearly all of the available bandwidth. You shouldn't be too surprised if other things suffer. NeilBrown Understood, will reduce this, thanks. Justin. - To unsubscribe from this list: send the line unsubscribe linux-raid in the body of a message to [EMAIL PROTECTED] More majordomo info at http://vger.kernel.org/majordomo-info.html
Re: is this raid5 OK ?
hi, 1) the kernel was: [EMAIL PROTECTED] ~]# uname -a Linux alfred 2.6.19-1.2288.fc5xen0 #1 SMP Sat Feb 10 16:57:02 EST 2007 i686 athlon i386 GNU/Linux now upgraded to: [EMAIL PROTECTED] ~]# uname -a Linux alfred 2.6.20-1.2307.fc5xen0 #1 SMP Sun Mar 18 21:59:42 EDT 2007 i686 athlon i386 GNU/Linux OS is fedora core 6 [EMAIL PROTECTED] ~]# mdadm --version mdadm - v2.3.1 - 6 February 2006 2) I got the impression that the old 350W power supply was to weak, I replaced it by a 400W version. 3) re-created the raid: [EMAIL PROTECTED] ~]# mdadm --misc --zero-superblock /dev/hde1 [EMAIL PROTECTED] ~]# mdadm --misc --zero-superblock /dev/hdf1 [EMAIL PROTECTED] ~]# mdadm --misc --zero-superblock /dev/hdg1 [EMAIL PROTECTED] ~]# mdadm --misc --zero-superblock /dev/hdh1 [EMAIL PROTECTED] ~]# mdadm --create --verbose /dev/md0 --level=5 --raid-devices=4 --spare-devices=0 /dev/hde1 /dev/hdf1 /dev/hdg1 /dev/hdh1 mdadm: layout defaults to left-symmetric mdadm: chunk size defaults to 64K mdadm: size set to 390708736K mdadm: array /dev/md0 started. [EMAIL PROTECTED] ~]# cat /proc/mdstat Personalities : [raid6] [raid5] [raid4] md0 : active raid5 hdh1[4] hdg1[2] hdf1[1] hde1[0] 1172126208 blocks level 5, 64k chunk, algorithm 2 [4/3] [UUU_] unused devices: none same as before. 4) did as dan suggested: [EMAIL PROTECTED] ~]# mdadm -S /dev/md0 [EMAIL PROTECTED] ~]# mdadm --misc --zero-superblock /dev/hde1 [EMAIL PROTECTED] ~]# mdadm --misc --zero-superblock /dev/hdf1 [EMAIL PROTECTED] ~]# mdadm --misc --zero-superblock /dev/hdg1 [EMAIL PROTECTED] ~]# mdadm --misc --zero-superblock /dev/hdh1 [EMAIL PROTECTED] ~]# mdadm --create /dev/md0 -n 4 -l 5 /dev/hd[efg]1 missing mdadm: array /dev/md0 started. [EMAIL PROTECTED] ~]# cat /proc/mdstat Personalities : [raid6] [raid5] [raid4] md0 : active raid5 hdg1[2] hdf1[1] hde1[0] 1172126208 blocks level 5, 64k chunk, algorithm 2 [4/3] [UUU_] unused devices: none [EMAIL PROTECTED] ~]# mdadm --add /dev/md0 /dev/hdh1 mdadm: added /dev/hdh1 [EMAIL PROTECTED] ~]# cat /proc/mdstat Personalities : [raid6] [raid5] [raid4] md0 : active raid5 hdh1[4] hdg1[2] hdf1[1] hde1[0] 1172126208 blocks level 5, 64k chunk, algorithm 2 [4/3] [UUU_] [] recovery = 0.0% (47984/390708736) finish=406.9min speed=15994K/sec unused devices: none seems like it's working now - tnx ! cu - To unsubscribe from this list: send the line unsubscribe linux-raid in the body of a message to [EMAIL PROTECTED] More majordomo info at http://vger.kernel.org/majordomo-info.html
Re: is this raid5 OK ?
Rainer Fuegenstein wrote: hi, 1) the kernel was: [EMAIL PROTECTED] ~]# uname -a Linux alfred 2.6.19-1.2288.fc5xen0 #1 SMP Sat Feb 10 16:57:02 EST 2007 i686 athlon i386 GNU/Linux now upgraded to: [EMAIL PROTECTED] ~]# uname -a Linux alfred 2.6.20-1.2307.fc5xen0 #1 SMP Sun Mar 18 21:59:42 EDT 2007 i686 athlon i386 GNU/Linux OS is fedora core 6 [EMAIL PROTECTED] ~]# mdadm --version mdadm - v2.3.1 - 6 February 2006 2) I got the impression that the old 350W power supply was to weak, I replaced it by a 400W version. 3) re-created the raid: [EMAIL PROTECTED] ~]# mdadm --misc --zero-superblock /dev/hde1 [EMAIL PROTECTED] ~]# mdadm --misc --zero-superblock /dev/hdf1 [EMAIL PROTECTED] ~]# mdadm --misc --zero-superblock /dev/hdg1 [EMAIL PROTECTED] ~]# mdadm --misc --zero-superblock /dev/hdh1 [EMAIL PROTECTED] ~]# mdadm --create --verbose /dev/md0 --level=5 --raid-devices=4 --spare-devices=0 /dev/hde1 /dev/hdf1 /dev/hdg1 /dev/hdh1 mdadm: layout defaults to left-symmetric mdadm: chunk size defaults to 64K mdadm: size set to 390708736K mdadm: array /dev/md0 started. [EMAIL PROTECTED] ~]# cat /proc/mdstat Personalities : [raid6] [raid5] [raid4] md0 : active raid5 hdh1[4] hdg1[2] hdf1[1] hde1[0] 1172126208 blocks level 5, 64k chunk, algorithm 2 [4/3] [UUU_] unused devices: none same as before. 4) did as dan suggested: [EMAIL PROTECTED] ~]# mdadm -S /dev/md0 [EMAIL PROTECTED] ~]# mdadm --misc --zero-superblock /dev/hde1 [EMAIL PROTECTED] ~]# mdadm --misc --zero-superblock /dev/hdf1 [EMAIL PROTECTED] ~]# mdadm --misc --zero-superblock /dev/hdg1 [EMAIL PROTECTED] ~]# mdadm --misc --zero-superblock /dev/hdh1 [EMAIL PROTECTED] ~]# mdadm --create /dev/md0 -n 4 -l 5 /dev/hd[efg]1 missing mdadm: array /dev/md0 started. [EMAIL PROTECTED] ~]# cat /proc/mdstat Personalities : [raid6] [raid5] [raid4] md0 : active raid5 hdg1[2] hdf1[1] hde1[0] 1172126208 blocks level 5, 64k chunk, algorithm 2 [4/3] [UUU_] unused devices: none [EMAIL PROTECTED] ~]# mdadm --add /dev/md0 /dev/hdh1 mdadm: added /dev/hdh1 [EMAIL PROTECTED] ~]# cat /proc/mdstat Personalities : [raid6] [raid5] [raid4] md0 : active raid5 hdh1[4] hdg1[2] hdf1[1] hde1[0] 1172126208 blocks level 5, 64k chunk, algorithm 2 [4/3] [UUU_] [] recovery = 0.0% (47984/390708736) finish=406.9min speed=15994K/sec unused devices: none seems like it's working now - tnx ! This still looks odd, why should it behave like this. I have created a lot of arrays (when I was doing the RAID5 speed testing thread), and never had anything like this. I'd like to see dmesg to see if there was an error reported regarding this. I think there's more going on, the original post showed the array as up rather than some building status, also indicates some issue, perhaps. What is the partition type of each of these partitions? Perhaps there's a clue there. -- bill davidsen [EMAIL PROTECTED] CTO TMR Associates, Inc Doing interesting things with small computers since 1979 - To unsubscribe from this list: send the line unsubscribe linux-raid in the body of a message to [EMAIL PROTECTED] More majordomo info at http://vger.kernel.org/majordomo-info.html
Re: mdadm: RUN_ARRAY failed: Cannot allocate memory
Neil Brown wrote: On Saturday March 24, [EMAIL PROTECTED] wrote: Hello Neil , I found the problem that caused the 'cannot allcate memory' , DON'T use '--bitmap=' . But that said , H , Shouldn't mdadm just stop say ... 'md: bitmaps not supported for this level.' Like it puts out into dmesg . Also think this message in dmesg is interesting . raid0: bad disk number -1 - aborting!' Hth , JimL Yeah mdadm should be fixed too, but this kernel patch should make it behave a bit better. I'll queue it for 2.6.22. Given the release cycle, this might fit 2.6.21-rc6 (is is a fix), or stable 2.6.21.1 if 21 comes out soon. In any case it could go in -mm for testing and to be sure it would be pushed at an appropriate time. Thanks, NeilBrown Move test for whether level supports bitmap to correct place. We need to check for internal-consistency of superblock in load_super. validate_super is for inter-device consistency. Signed-off-by: Neil Brown [EMAIL PROTECTED] ### Diffstat output ./drivers/md/md.c | 42 ++ 1 file changed, 26 insertions(+), 16 deletions(-) diff .prev/drivers/md/md.c ./drivers/md/md.c --- .prev/drivers/md/md.c 2007-03-29 16:42:18.0 +1000 +++ ./drivers/md/md.c 2007-03-29 16:49:26.0 +1000 @@ -695,6 +695,17 @@ static int super_90_load(mdk_rdev_t *rde rdev-data_offset = 0; rdev-sb_size = MD_SB_BYTES; + if (sb-state (1MD_SB_BITMAP_PRESENT)) { + if (sb-level != 1 sb-level != 4 +sb-level != 5 sb-level != 6 +sb-level != 10) { + /* FIXME use a better test */ + printk(KERN_WARNING + md: bitmaps not supported for this level.\n); + goto abort; + } + } + if (sb-level == LEVEL_MULTIPATH) rdev-desc_nr = -1; else @@ -793,16 +804,8 @@ static int super_90_validate(mddev_t *md mddev-max_disks = MD_SB_DISKS; if (sb-state (1MD_SB_BITMAP_PRESENT) - mddev-bitmap_file == NULL) { - if (mddev-level != 1 mddev-level != 4 -mddev-level != 5 mddev-level != 6 -mddev-level != 10) { - /* FIXME use a better test */ - printk(KERN_WARNING md: bitmaps not supported for this level.\n); - return -EINVAL; - } + mddev-bitmap_file == NULL) mddev-bitmap_offset = mddev-default_bitmap_offset; - } } else if (mddev-pers == NULL) { /* Insist on good event counter while assembling */ @@ -1059,6 +1062,18 @@ static int super_1_load(mdk_rdev_t *rdev bdevname(rdev-bdev,b)); return -EINVAL; } + if ((le32_to_cpu(sb-feature_map) MD_FEATURE_BITMAP_OFFSET)) { + if (sb-level != cpu_to_le32(1) + sb-level != cpu_to_le32(4) + sb-level != cpu_to_le32(5) + sb-level != cpu_to_le32(6) + sb-level != cpu_to_le32(10)) { + printk(KERN_WARNING + md: bitmaps not supported for this level.\n); + return -EINVAL; + } + } + rdev-preferred_minor = 0x; rdev-data_offset = le64_to_cpu(sb-data_offset); atomic_set(rdev-corrected_errors, le32_to_cpu(sb-cnt_corrected_read)); @@ -1142,14 +1157,9 @@ static int super_1_validate(mddev_t *mdd mddev-max_disks = (4096-256)/2; if ((le32_to_cpu(sb-feature_map) MD_FEATURE_BITMAP_OFFSET) - mddev-bitmap_file == NULL ) { - if (mddev-level != 1 mddev-level != 5 mddev-level != 6 -mddev-level != 10) { - printk(KERN_WARNING md: bitmaps not supported for this level.\n); - return -EINVAL; - } + mddev-bitmap_file == NULL ) mddev-bitmap_offset = (__s32)le32_to_cpu(sb-bitmap_offset); - } + if ((le32_to_cpu(sb-feature_map) MD_FEATURE_RESHAPE_ACTIVE)) { mddev-reshape_position = le64_to_cpu(sb-reshape_position); mddev-delta_disks = le32_to_cpu(sb-delta_disks); -- bill davidsen [EMAIL PROTECTED] CTO TMR Associates, Inc Doing interesting things with small computers since 1979 - To unsubscribe from this list: send the line unsubscribe linux-raid in the body of a message to [EMAIL PROTECTED] More majordomo info at http://vger.kernel.org/majordomo-info.html
Re: is this raid5 OK ?
Bill Davidsen wrote: This still looks odd, why should it behave like this. I have created a lot of arrays (when I was doing the RAID5 speed testing thread), and never had anything like this. I'd like to see dmesg to see if there was an error reported regarding this. I think there's more going on, the original post showed the array as up rather than some building status, also indicates some issue, perhaps. What is the partition type of each of these partitions? Perhaps there's a clue there. partition type is FD (linux raid autodetect) on all disks. here's some more info: the hardware is pretty old, an 800MHz ASUS board with AMD cpu and an extra onboard promise IDE controller with two channels. the server was working well with a 60 GB hda disk (system) and a single 400 GB disk (hde) for data. kernel was 2.6.19-1.2288.fc5xen0. when I added 3 more 400 GB disks (hdf to hdh) and created the raid5, the server crashed (rebooted, freezed, ...) as soon as there was more activity on the raid (kernel panics indicating trouble with interrupts, inpage errors etc.) I then upgraded to a 400W power supply, which didn't help. I went back to two single (non-raid) 400 GB disks - same problem. finally, I figured out that the non-xen kernel works without problems. I'm filling the raid5 since several hours now and the system is still stable. I haven't tried to re-create the raid5 using the non-xen kernel, it was created using the xen kernel. maybe xen could be the problem ? I was wrong in my last post - OS is actually fedora core 5 (sorry for the typo) current state of the raid5: [EMAIL PROTECTED] ~]# mdadm --detail --scan ARRAY /dev/md0 level=raid5 num-devices=4 spares=1 UUID=e96cd8fe:c56c3438:6d9b6c14:9f0eebda [EMAIL PROTECTED] ~]# mdadm --misc --detail /dev/md0 /dev/md0: Version : 00.90.03 Creation Time : Fri Mar 30 15:55:42 2007 Raid Level : raid5 Array Size : 1172126208 (1117.83 GiB 1200.26 GB) Device Size : 390708736 (372.61 GiB 400.09 GB) Raid Devices : 4 Total Devices : 4 Preferred Minor : 0 Persistence : Superblock is persistent Update Time : Fri Mar 30 20:22:27 2007 State : active, degraded, recovering Active Devices : 3 Working Devices : 4 Failed Devices : 0 Spare Devices : 1 Layout : left-symmetric Chunk Size : 64K Rebuild Status : 12% complete UUID : e96cd8fe:c56c3438:6d9b6c14:9f0eebda Events : 0.26067 Number Major Minor RaidDevice State 0 3310 active sync /dev/hde1 1 33 651 active sync /dev/hdf1 2 3412 active sync /dev/hdg1 4 34 653 spare rebuilding /dev/hdh1 here's the dmesg of the last reboot (when the raid was already created, but still syncing): Linux version 2.6.20-1.2307.fc5 ([EMAIL PROTECTED]) (gcc version 4.1.1 20070105 (Red Hat 4.1.1-51)) #1 Sun Mar 18 20:44:48 EDT 2007 BIOS-provided physical RAM map: sanitize start sanitize end copy_e820_map() start: size: 0009f000 end: 0009f000 type: 1 copy_e820_map() type is E820_RAM copy_e820_map() start: 0009f000 size: 1000 end: 000a type: 2 copy_e820_map() start: 000f size: 0001 end: 0010 type: 2 copy_e820_map() start: 0010 size: 1feec000 end: 1ffec000 type: 1 copy_e820_map() type is E820_RAM copy_e820_map() start: 1ffec000 size: 3000 end: 1ffef000 type: 3 copy_e820_map() start: 1ffef000 size: 0001 end: 1000 type: 2 copy_e820_map() start: 1000 size: 1000 end: 2000 type: 4 copy_e820_map() start: size: 0001 end: 0001 type: 2 BIOS-e820: - 0009f000 (usable) BIOS-e820: 0009f000 - 000a (reserved) BIOS-e820: 000f - 0010 (reserved) BIOS-e820: 0010 - 1ffec000 (usable) BIOS-e820: 1ffec000 - 1ffef000 (ACPI data) BIOS-e820: 1ffef000 - 1000 (reserved) BIOS-e820: 1000 - 2000 (ACPI NVS) BIOS-e820: - 0001 (reserved) 0MB HIGHMEM available. 511MB LOWMEM available. Using x86 segment limits to approximate NX protection Entering add_active_range(0, 0, 131052) 0 entries of 256 used Zone PFN ranges: DMA 0 - 4096 Normal 4096 - 131052 HighMem131052 - 131052 early_node_map[1] active PFN ranges 0:0 - 131052 On node 0 totalpages: 131052 DMA zone: 32 pages used for memmap DMA zone: 0 pages reserved DMA zone: 4064 pages, LIFO batch:0 Normal zone: 991 pages used for memmap Normal zone: 125965 pages, LIFO batch:31 HighMem zone: 0 pages used for memmap DMI 2.3 present. Using APIC driver default ACPI: RSDP
Re: is this raid5 OK ?
On Fri, 30 Mar 2007, Rainer Fuegenstein wrote: Bill Davidsen wrote: This still looks odd, why should it behave like this. I have created a lot of arrays (when I was doing the RAID5 speed testing thread), and never had anything like this. I'd like to see dmesg to see if there was an error reported regarding this. I think there's more going on, the original post showed the array as up rather than some building status, also indicates some issue, perhaps. What is the partition type of each of these partitions? Perhaps there's a clue there. partition type is FD (linux raid autodetect) on all disks. here's some more info: the hardware is pretty old, an 800MHz ASUS board with AMD cpu and an extra onboard promise IDE controller with two channels. the server was working well with a 60 GB hda disk (system) and a single 400 GB disk (hde) for data. kernel was 2.6.19-1.2288.fc5xen0. when I added 3 more 400 GB disks (hdf to hdh) and created the raid5, the server crashed (rebooted, freezed, ...) as soon as there was more activity on the raid (kernel panics indicating trouble with interrupts, inpage errors etc.) I then upgraded to a 400W power supply, which didn't help. I went back to two single (non-raid) 400 GB disks - same problem. finally, I figured out that the non-xen kernel works without problems. I'm filling the raid5 since several hours now and the system is still stable. I haven't tried to re-create the raid5 using the non-xen kernel, it was created using the xen kernel. maybe xen could be the problem ? I was wrong in my last post - OS is actually fedora core 5 (sorry for the typo) PCI: Disabling Via external APIC routing I will note there is the ominous '400GB' lockup bug with certain promise controllers. With the Promise ATA/133 controllers in some configurations you will get a DRQ/lockup no matter what, replacing with an ATA/100 card and no issues. But I see you have a 20265 with is an ATA/100 promise/chipset. Just out of curiosity have you tried writing or running badblocks on each parition simultaenously, this would simulate (somewhat) the I/O sent/received to the drives during a RAID5 rebuild. Justin. - To unsubscribe from this list: send the line unsubscribe linux-raid in the body of a message to [EMAIL PROTECTED] More majordomo info at http://vger.kernel.org/majordomo-info.html
Re: raid5 write performance
Please see bellow. On 8/28/06, Neil Brown [EMAIL PROTECTED] wrote: On Sunday August 13, [EMAIL PROTECTED] wrote: well ... me again Following your advice I added a deadline for every WRITE stripe head when it is created. in raid5_activate_delayed i checked if deadline is expired and if not i am setting the sh to prereadactive mode as . This small fix ( and in few other places in the code) reduced the amount of reads to zero with dd but with no improvement to throghput. But with random access to the raid ( buffers are aligned by the stripe width and with the size of stripe width ) there is an improvement of at least 20 % . Problem is that a user must know what he is doing else there would be a reduction in performance if deadline line it too long (say 100 ms). So if I understand you correctly, you are delaying write requests to partial stripes slightly (your 'deadline') and this is sometimes giving you a 20% improvement ? I'm not surprised that you could get some improvement. 20% is quite surprising. It would be worth following through with this to make that improvement generally available. As you say, picking a time in milliseconds is very error prone. We really need to come up with something more natural. I had hopped that the 'unplug' infrastructure would provide the right thing, but apparently not. Maybe unplug is just being called too often. I'll see if I can duplicate this myself and find out what is really going on. Thanks for the report. NeilBrown Neil Hello. I am sorry for this interval , I was assigned abruptly to a different project. 1. I'd taken a look at the raid5 delay patch I have written a while ago. I ported it to 2.6.17 and tested it. it makes sounds of working and when used correctly it eliminates the reads penalty. 2. Benchmarks . configuration: I am testing a raid5 x 3 disks with 1MB chunk size. IOs are synchronous and non-buffered(o_direct) , 2 MB in size and always aligned to the beginning of a stripe. kernel is 2.6.17. The stripe_delay was set to 10ms. Attached is the simple_write code. command : simple_write /dev/md1 2048 0 1000 simple_write raw writes (O_DIRECT) sequentially starting from offset zero 2048 kilobytes 1000 times. Benchmark Before patch sda1848.00 8384.00 50992.00 8384 50992 sdb1995.00 12424.00 51008.00 12424 51008 sdc1698.00 8160.00 51000.00 8160 51000 sdd 0.00 0.00 0.00 0 0 md0 0.00 0.00 0.00 0 0 md1 450.00 0.00102400.00 0 102400 Benchmark After patch sda 389.11 0.00128530.69 0 129816 sdb 381.19 0.00129354.46 0 130648 sdc 383.17 0.00128530.69 0 129816 sdd 0.00 0.00 0.00 0 0 md0 0.00 0.00 0.00 0 0 md11140.59 0.00259548.51 0 262144 As one can see , no additional reads were done. One can actually calculate the raid's utilization: n-1/n * ( single disk throughput with 1M writes ) . 3. The patch code. Kernel tested above was 2.6.17. The patch is of 2.6.20.2 because I have noticed a big code differences between 17 to 20.x . This patch was not tested on 2.6.20.2 but it is essentialy the same. I have not tested (yet) degraded mode or any other non-common pathes. --- linux-2.6.20.2/drivers/md/raid5.c 2007-03-09 20:58:04.0 +0200 +++ linux-2.6.20.2-raid/drivers/md/raid5.c 2007-03-30 12:37:55.0 +0300 @@ -65,6 +65,7 @@ #define NR_HASH(PAGE_SIZE / sizeof(struct hlist_head)) #define HASH_MASK (NR_HASH - 1) + #define stripe_hash(conf, sect) (((conf)-stripe_hashtbl[((sect) STRIPE_SHIFT) HASH_MASK])) /* bio's attached to a stripe+device for I/O are linked together in bi_sector @@ -234,6 +235,8 @@ sh-sector = sector; sh-pd_idx = pd_idx; sh-state = 0; + sh-active_preread_jiffies = + msecs_to_jiffies( atomic_read(conf-deadline_ms) )+ jiffies; sh-disks = disks; @@ -628,6 +631,7 @@ clear_bit(R5_LOCKED, sh-dev[i].flags); set_bit(STRIPE_HANDLE, sh-state); + sh-active_preread_jiffies = jiffies; release_stripe(sh); return 0; } @@ -1255,8 +1259,11 @@ bip = sh-dev[dd_idx].towrite; if (*bip == NULL sh-dev[dd_idx].written == NULL) firstwrite = 1; - } else + } else{ bip = sh-dev[dd_idx].toread; + sh-active_preread_jiffies = jiffies; + } + while (*bip (*bip)-bi_sector bi-bi_sector) { if ((*bip)-bi_sector + ((*bip)-bi_size 9) bi-bi_sector)
Re: is this raid5 OK ?
-wheneverRainer Fuegenstein wrote: Bill Davidsen wrote: This still looks odd, why should it behave like this. I have created a lot of arrays (when I was doing the RAID5 speed testing thread), and never had anything like this. I'd like to see dmesg to see if there was an error reported regarding this. I think there's more going on, the original post showed the array as up rather than some building status, also indicates some issue, perhaps. What is the partition type of each of these partitions? Perhaps there's a clue there. partition type is FD (linux raid autodetect) on all disks. here's some more info: the hardware is pretty old, an 800MHz ASUS board with AMD cpu and an extra onboard promise IDE controller with two channels. the server was working well with a 60 GB hda disk (system) and a single 400 GB disk (hde) for data. kernel was 2.6.19-1.2288.fc5xen0. when I added 3 more 400 GB disks (hdf to hdh) and created the raid5, the server crashed (rebooted, freezed, ...) as soon as there was more activity on the raid (kernel panics indicating trouble with interrupts, inpage errors etc.) I then upgraded to a 400W power supply, which didn't help. I went back to two single (non-raid) 400 GB disks - same problem. finally, I figured out that the non-xen kernel works without problems. I'm filling the raid5 since several hours now and the system is still stable. I haven't tried to re-create the raid5 using the non-xen kernel, it was created using the xen kernel. maybe xen could be the problem ? I think it sounds likely at this point, I have been having issues with xen FC6 kernels, so perhaps the build or testing environment has changed. However, I would round up the usual suspects, check all cables tight, check master/slave jumper settings on drives, etc. Be sure you have the appropriate cables, 80 pin where needed. Unless you need the xen kernel you might be better off without it for now. The rest of your details were complete but didn't give me a clue, sorry. I was wrong in my last post - OS is actually fedora core 5 (sorry for the typo) current state of the raid5: [EMAIL PROTECTED] ~]# mdadm --detail --scan ARRAY /dev/md0 level=raid5 num-devices=4 spares=1 UUID=e96cd8fe:c56c3438:6d9b6c14:9f0eebda [EMAIL PROTECTED] ~]# mdadm --misc --detail /dev/md0 /dev/md0: Version : 00.90.03 Creation Time : Fri Mar 30 15:55:42 2007 Raid Level : raid5 Array Size : 1172126208 (1117.83 GiB 1200.26 GB) Device Size : 390708736 (372.61 GiB 400.09 GB) Raid Devices : 4 Total Devices : 4 Preferred Minor : 0 Persistence : Superblock is persistent Update Time : Fri Mar 30 20:22:27 2007 State : active, degraded, recovering Active Devices : 3 Working Devices : 4 Failed Devices : 0 Spare Devices : 1 Layout : left-symmetric Chunk Size : 64K Rebuild Status : 12% complete UUID : e96cd8fe:c56c3438:6d9b6c14:9f0eebda Events : 0.26067 Number Major Minor RaidDevice State 0 3310 active sync /dev/hde1 1 33 651 active sync /dev/hdf1 2 3412 active sync /dev/hdg1 4 34 653 spare rebuilding /dev/hdh1 here's the dmesg of the last reboot (when the raid was already created, but still syncing): [ since it told me nothing useful I deleted it ] -- bill davidsen [EMAIL PROTECTED] CTO TMR Associates, Inc Doing interesting things with small computers since 1979 - To unsubscribe from this list: send the line unsubscribe linux-raid in the body of a message to [EMAIL PROTECTED] More majordomo info at http://vger.kernel.org/majordomo-info.html
Re: is this raid5 OK ?
Justin Piszcz wrote: On Fri, 30 Mar 2007, Rainer Fuegenstein wrote: Bill Davidsen wrote: This still looks odd, why should it behave like this. I have created a lot of arrays (when I was doing the RAID5 speed testing thread), and never had anything like this. I'd like to see dmesg to see if there was an error reported regarding this. I think there's more going on, the original post showed the array as up rather than some building status, also indicates some issue, perhaps. What is the partition type of each of these partitions? Perhaps there's a clue there. partition type is FD (linux raid autodetect) on all disks. here's some more info: the hardware is pretty old, an 800MHz ASUS board with AMD cpu and an extra onboard promise IDE controller with two channels. the server was working well with a 60 GB hda disk (system) and a single 400 GB disk (hde) for data. kernel was 2.6.19-1.2288.fc5xen0. when I added 3 more 400 GB disks (hdf to hdh) and created the raid5, the server crashed (rebooted, freezed, ...) as soon as there was more activity on the raid (kernel panics indicating trouble with interrupts, inpage errors etc.) I then upgraded to a 400W power supply, which didn't help. I went back to two single (non-raid) 400 GB disks - same problem. finally, I figured out that the non-xen kernel works without problems. I'm filling the raid5 since several hours now and the system is still stable. I haven't tried to re-create the raid5 using the non-xen kernel, it was created using the xen kernel. maybe xen could be the problem ? I was wrong in my last post - OS is actually fedora core 5 (sorry for the typo) PCI: Disabling Via external APIC routing I will note there is the ominous '400GB' lockup bug with certain promise controllers. With the Promise ATA/133 controllers in some configurations you will get a DRQ/lockup no matter what, replacing with an ATA/100 card and no issues. But I see you have a 20265 with is an ATA/100 promise/chipset. Just out of curiosity have you tried writing or running badblocks on each parition simultaenously, this would simulate (somewhat) the I/O sent/received to the drives during a RAID5 rebuild. These are all things which could be related, but any clue why the non-xen kernel works? -- bill davidsen [EMAIL PROTECTED] CTO TMR Associates, Inc Doing interesting things with small computers since 1979 - To unsubscribe from this list: send the line unsubscribe linux-raid in the body of a message to [EMAIL PROTECTED] More majordomo info at http://vger.kernel.org/majordomo-info.html