Erreur disque

yfi · Décembre 14, 2016, 9:14am

Bonjour,

Il me semble que les disques d’un serveur commencent de devenir vieux, dans les logs je voix les erreurs suivantes :

Dec 13 23:02:37 server kernel: [15646652.731034] attempt to access beyond end of device
Dec 13 23:02:37 server kernel: [15646652.737136] md1: rw=529, want=486026664, limit=486026624
Dec 13 23:02:37 server kernel: [15646652.743135] Buffer I/O error on device md1, logical block 60753332
Dec 13 23:02:37 server kernel: [15646652.749033] lost page write due to I/O error on md1
Dec 13 23:02:37 server kernel: [15646652.754827] attempt to access beyond end of device
Dec 13 23:02:37 server kernel: [15646652.760642] md1: rw=529, want=486026672, limit=486026624
Dec 13 23:02:37 server kernel: [15646652.766470] Buffer I/O error on device md1, logical block 60753333
Dec 13 23:02:37 server kernel: [15646652.772210] lost page write due to I/O error on md1
Dec 13 23:02:37 server kernel: [15646652.777872] attempt to access beyond end of device
Dec 13 23:02:37 server kernel: [15646652.783480] md1: rw=529, want=486026680, limit=486026624
Dec 13 23:02:37 server kernel: [15646652.789101] Buffer I/O error on device md1, logical block 60753334
Dec 13 23:02:37 server kernel: [15646652.794742] lost page write due to I/O error on md1
Dec 13 23:02:37 server kernel: [15646652.800307] attempt to access beyond end of device
Dec 13 23:02:37 server kernel: [15646652.805795] md1: rw=529, want=486026688, limit=486026624
Dec 13 23:02:37 server kernel: [15646652.811222] Buffer I/O error on device md1, logical block 60753335
Dec 13 23:02:37 server kernel: [15646652.816647] lost page write due to I/O error on md1
Dec 13 23:02:37 server kernel: [15646652.822108] attempt to access beyond end of device
Dec 13 23:02:37 server kernel: [15646652.827569] md1: rw=529, want=486026696, limit=486026624
Dec 13 23:02:37 server kernel: [15646652.832958] Buffer I/O error on device md1, logical block 60753336
Dec 13 23:02:37 server kernel: [15646652.838251] lost page write due to I/O error on md1
Dec 13 23:02:37 server kernel: [15646652.843538] attempt to access beyond end of device
Dec 13 23:02:37 server kernel: [15646652.848802] md1: rw=529, want=486026704, limit=486026624
Dec 13 23:02:37 server kernel: [15646652.854094] Buffer I/O error on device md1, logical block 60753337
Dec 13 23:02:37 server kernel: [15646652.859302] lost page write due to I/O error on md1
Dec 13 23:02:37 server kernel: [15646652.864478] attempt to access beyond end of device
Dec 13 23:02:37 server kernel: [15646652.869570] md1: rw=529, want=486026712, limit=486026624
Dec 13 23:02:37 server kernel: [15646652.874672] attempt to access beyond end of device
Dec 13 23:02:37 server kernel: [15646652.879764] md1: rw=529, want=486026720, limit=486026624
Dec 13 23:02:37 server kernel: [15646652.884890] attempt to access beyond end of device
Dec 13 23:02:37 server kernel: [15646652.889907] md1: rw=529, want=486026728, limit=486026624
Dec 13 23:02:37 server kernel: [15646652.894829] attempt to access beyond end of device
Dec 13 23:02:37 server kernel: [15646652.899702] md1: rw=529, want=486026736, limit=486026624
Dec 13 23:02:37 server kernel: [15646652.904577] attempt to access beyond end of device
Dec 13 23:02:37 server kernel: [15646652.909442] md1: rw=529, want=486026744, limit=486026624
Dec 13 23:02:37 server kernel: [15646652.914321] attempt to access beyond end of device
Dec 13 23:02:37 server kernel: [15646652.919153] md1: rw=529, want=486026752, limit=486026624
Dec 13 23:02:37 server kernel: [15646652.924020] attempt to access beyond end of device
Dec 13 23:02:37 server kernel: [15646652.928897] md1: rw=529, want=486026760, limit=486026624
Dec 13 23:02:37 server kernel: [15646652.933771] attempt to access beyond end of device
Dec 13 23:02:37 server kernel: [15646652.938594] md1: rw=529, want=486026768, limit=486026624
Dec 13 23:02:37 server kernel: [15646652.943462] attempt to access beyond end of device
Dec 13 23:02:37 server kernel: [15646652.948310] md1: rw=529, want=486026776, limit=486026624
Dec 13 23:02:37 server kernel: [15646652.953179] attempt to access beyond end of device
Dec 13 23:02:37 server kernel: [15646652.957984] md1: rw=529, want=486026784, limit=486026624
Dec 13 23:02:37 server kernel: [15646652.962827] attempt to access beyond end of device
Dec 13 23:02:37 server kernel: [15646652.967652] md1: rw=529, want=486026792, limit=486026624
Dec 13 23:02:37 server kernel: [15646652.972511] attempt to access beyond end of device
Dec 13 23:02:37 server kernel: [15646652.977314] md1: rw=529, want=486026800, limit=486026624
Dec 13 23:02:37 server kernel: [15646652.982161] attempt to access beyond end of device
Dec 13 23:02:37 server kernel: [15646652.986978] md1: rw=529, want=486026808, limit=486026624
Dec 13 23:02:37 server kernel: [15646653.239790] JBD: Detected IO errors while flushing file data on md1

Je n’est pas encore lancé de contrôle SMART j’attends la fin de journée pour ne pas impacter les perf de serveur de prod, pouvez vous me dire d’où peuvent venir ces erreurs ?

Merci par avance.

PascalHambourg · Décembre 14, 2016, 11:33am

La lecture des données SMART d’un disque avec smartctl ou autre n’impacte pas les performances, ce n’est que de la lecture d’informations déjà emmagasinées par le contrôleur intégré au disque.

Mais je ne pense pas que ce soit une erreur physique. Les messages disent qu’il y a eu une tentative de lecture légèrement au delà de la fin du périphérique. Est-ce qu’un disque, une partition, un ensemble RAID aurait été redimensionné ?
Informations utiles :

tune2fs -l /dev/md1
cat /proc/mdstat
mdadm --detail /dev/md1
mdadm --examine /dev/sd* # ne mettre que les partitions/disques membres de md1
fdisk -l /dev/sd* # pour chaque disque supportant un membre de md1

yfi · Décembre 14, 2016, 9:59am

Non pas de d’intervention sur les disques.

fdisk -l

Disk /dev/sdb: 250.0 GB, 250000000000 bytes
255 heads, 63 sectors/track, 30394 cylinders, total 488281250 sectors
Units = sectors of 1 * 512 = 512 bytes
Sector size (logical/physical): 512 bytes / 512 bytes
I/O size (minimum/optimal): 512 bytes / 512 bytes
Disk identifier: 0x00000080

Device Boot Start End Blocks Id System
/dev/sdb1 * 1 204799 102399+ 83 Linux
/dev/sdb2 204800 2252799 1024000 82 Linux swap / Solaris
/dev/sdb3 2252800 488279609 243013405 83 Linux

Disk /dev/sda: 250.0 GB, 250000000000 bytes
255 heads, 63 sectors/track, 30394 cylinders, total 488281250 sectors
Units = sectors of 1 * 512 = 512 bytes
Sector size (logical/physical): 512 bytes / 512 bytes
I/O size (minimum/optimal): 512 bytes / 512 bytes
Disk identifier: 0x00092940

Device Boot Start End Blocks Id System
/dev/sda1 * 1 204799 102399+ 83 Linux
/dev/sda2 204800 2252799 1024000 82 Linux swap / Solaris
/dev/sda3 2252800 488279609 243013405 83 Linux

Disk /dev/md1: 248.8 GB, 248845631488 bytes
2 heads, 4 sectors/track, 60753328 cylinders, total 486026624 sectors
Units = sectors of 1 * 512 = 512 bytes
Sector size (logical/physical): 512 bytes / 512 bytes
I/O size (minimum/optimal): 512 bytes / 512 bytes
Disk identifier: 0x00000000

Disk /dev/md1 doesn’t contain a valid partition table

Disk /dev/md0: 104 MB, 104726528 bytes
2 heads, 4 sectors/track, 25568 cylinders, total 204544 sectors
Units = sectors of 1 * 512 = 512 bytes
Sector size (logical/physical): 512 bytes / 512 bytes
I/O size (minimum/optimal): 512 bytes / 512 bytes
Disk identifier: 0x00000000

Disk /dev/md0 doesn’t contain a valid partition table

cat /proc/mdstat

Personalities : [raid1]
md0 : active raid1 sda1[0] sdb1[1]
102272 blocks [2/2] [UU]

md1 : active raid1 sda3[0] sdb3[1]
243013312 blocks [2/2] [UU]

unused devices:

mdadm -D /dev/md1

/dev/md1:
Version : 0.90
Creation Time : Mon Dec 3 17:05:07 2012
Raid Level : raid1
Array Size : 243013312 (231.76 GiB 248.85 GB)
Used Dev Size : 243013312 (231.76 GiB 248.85 GB)
Raid Devices : 2
Total Devices : 2
Preferred Minor : 1
Persistence : Superblock is persistent
Update Time : Wed Dec 14 10:39:31 2016
      State : clean
Active Devices : 2
Working Devices : 2
Failed Devices : 0
Spare Devices : 0
       UUID : 72413b9b:190cce1f:daf46747:b8a20270
     Events : 0.1601

Number   Major   Minor   RaidDevice State
   0       8        3        0      active sync   /dev/sda3
   1       8       19        1      active sync   /dev/sdb3
cat /proc/mdstat

Personalities : [raid1]
md0 : active raid1 sda1[0] sdb1[1]
102272 blocks [2/2] [UU]

md1 : active raid1 sda3[0] sdb3[1]
243013312 blocks [2/2] [UU]

unused devices:

sfdisk -l /dev/sda

Disk /dev/sda: 30394 cylinders, 255 heads, 63 sectors/track
Units = cylinders of 8225280 bytes, blocks of 1024 bytes, counting from 0

Device Boot Start End #cyls #blocks Id System
/dev/sda1 * 0+ 12- 13- 102399+ 83 Linux
/dev/sda2 12+ 140- 128- 1024000 82 Linux swap / Solaris
/dev/sda3 140+ 30393 30254- 243013405 83 Linux
/dev/sda4 0 - 0 0 0 Empty

sfdisk -l /dev/sdb

Disk /dev/sdb: 30394 cylinders, 255 heads, 63 sectors/track
Units = cylinders of 8225280 bytes, blocks of 1024 bytes, counting from 0

Device Boot Start End #cyls #blocks Id System
/dev/sdb1 * 0+ 12- 13- 102399+ 83 Linux
/dev/sdb2 12+ 140- 128- 1024000 82 Linux swap / Solaris
/dev/sdb3 140+ 30393 30254- 243013405 83 Linux
/dev/sdb4 0 - 0 0 0 Empty

PascalHambourg · Décembre 14, 2016, 11:39am

Apparemment c’est le système de fichiers (ext4 ?) qui essaie d’accéder au delà de la fin de /dev/md1.
En complément, qu’affichent

mdadm --examine /dev/sd[ab]3
tune2fs -l /dev/md1

PS : Pourquoi le swap n’est pas en RAID 1 ? En cas de défaillance d’un disque cela risque de causer un plantage du système.

GOGI · Décembre 14, 2016, 2:34pm

C’est la même personne à laquelle il avait été conseillé de mettre le swap en RAID1 il y a quelques jours dans un autre sujet.