More about: metadevice failed: invoke metareplace

From: Jordi Vidal (jordivi_at_wtransnet.net)
Date: 07/23/03

Next message: rich p: "jumpstart cdrom and nfs"

Previous message: Glenn, May (COMFIN, VFS, Consultant): "SUMMARY: Prom Password Lock"
Messages sorted by: [ date ] [ thread ] [ subject ] [ author ] [ attachment ]

Date: Wed, 23 Jul 2003 19:34:52 +0200 (CEST)
To: <sunmanagers@sunmanagers.org>

Thank you for your replies

I've been told to invoke "metareplace -e d1 c0t1d0s0" to try if
it fixes itself. No luck:

root@# metareplace -e d1 c0t1d0s0
metareplace: wtn450: d1: unknown metadevice type

I've been told also to run "metareplace -e d3 c0t1d0s0" but I
asume it is a mistake (isnt it?. The mistake is my poor knowledge?)

So I will try to format the drive and fmthard, installboot,
metadb, metareplace as it was a new brand drive. My question now is: Have
I to metadetach/metaclear the failed device d1 or it is destroyed yet?
(Can I run format right now or must I meta-something the bad drive?)

Regards,
Jordiv

-----------------------------------------------------------------------
Original Post:
Hi
I have a failed metadevice, "d1" which is a submirror. Reading
syslog I think there is a single bad block in the disk (Error Block:
1715719). My question is, can I reuse the device? marking bad blocks if
it is possible or should I remove and destroy the failed disk. I have
not any disk available to metareplace the device right now.

Instead of metareplace, can I safely "metadettach" and "metaclear"
the failed metadevice?

My first thought is to metadetach/metaclear all devices in the
failed disk (c0t1d0), wait for a new disk, reboot and recreate the mirror.
Any alternative?

d1: Submirror of d3
    State: Needs maintenance
    Invoke: metareplace d3 c0t1d0s0 <new device>
    Size: 66846720 blocks
    Stripe 0:
        Device Start Block Dbase State Hot Spare
        c0t1d0s0 0 No Maintenance

This is from my syslog:

Jul 22 08:37:14 wtn450 scsi: [ID 107833 kern.warning] WARNING: /pci@1f,4000/scsi@3/sd@1,0 (sd1):
Jul 22 08:37:14 wtn450 Error for Command: read(10) Error Level: Retryable
Jul 22 08:37:14 wtn450 scsi: [ID 107833 kern.notice] Requested Block: 1715568 Error Block: 1715719
Jul 22 08:37:14 wtn450 scsi: [ID 107833 kern.notice] Vendor: FUJITSU Serial Number: 0132M62300
Jul 22 08:37:14 wtn450 scsi: [ID 107833 kern.notice] Sense Key: Media Error
Jul 22 08:37:14 wtn450 scsi: [ID 107833 kern.notice] ASC: 0x11 (<vendor unique code 0x11>), ASCQ: 0x1, FRU: 0x0
Jul 22 08:37:15 wtn450 scsi: [ID 107833 kern.warning] WARNING: /pci@1f,4000/scsi@3/sd@1,0 (sd1):
Jul 22 08:37:15 wtn450 Error for Command: read(10) Error Level: Retryable
Jul 22 08:37:15 wtn450 scsi: [ID 107833 kern.notice] Requested Block: 1715568 Error Block: 1715719
Jul 22 08:37:15 wtn450 scsi: [ID 107833 kern.notice] Vendor: FUJITSU Serial Number: 0132M62300
Jul 22 08:37:15 wtn450 scsi: [ID 107833 kern.notice] Sense Key: Media Error
Jul 22 08:37:15 wtn450 scsi: [ID 107833 kern.notice] ASC: 0x11 (<vendor unique code 0x11>), ASCQ: 0x1, FRU: 0x0
Jul 22 08:37:16 wtn450 scsi: [ID 107833 kern.warning] WARNING: /pci@1f,4000/scsi@3/sd@1,0 (sd1):
Jul 22 08:37:16 wtn450 Error for Command: read(10) Error Level: Retryable
Jul 22 08:37:16 wtn450 scsi: [ID 107833 kern.notice] Requested Block: 1715568 Error Block: 1715718
Jul 22 08:37:16 wtn450 scsi: [ID 107833 kern.notice] Vendor: FUJITSU Serial Number: 0132M62300
Jul 22 08:37:16 wtn450 scsi: [ID 107833 kern.notice] Sense Key: Media Error
Jul 22 08:37:16 wtn450 scsi: [ID 107833 kern.notice] ASC: 0x11 (<vendor unique code 0x11>), ASCQ: 0x1, FRU: 0x0
Jul 22 08:37:16 wtn450 scsi: [ID 107833 kern.warning] WARNING: /pci@1f,4000/scsi@3/sd@1,0 (sd1):
Jul 22 08:37:16 wtn450 SCSI transport failed: reason 'reset': retrying command
Jul 22 08:37:19 wtn450 scsi: [ID 107833 kern.warning] WARNING: /pci@1f,4000/scsi@3/sd@1,0 (sd1):
Jul 22 08:37:19 wtn450 Error for Command: read(10) Error Level: Retryable
Jul 22 08:37:19 wtn450 scsi: [ID 107833 kern.notice] Requested Block: 1715568 Error Block: 1715719
Jul 22 08:37:19 wtn450 scsi: [ID 107833 kern.notice] Vendor: FUJITSU Serial Number: 0132M62300
Jul 22 08:37:19 wtn450 scsi: [ID 107833 kern.notice] Sense Key: Media Error
Jul 22 08:37:19 wtn450 scsi: [ID 107833 kern.notice] ASC: 0x11 (<vendor unique code 0x11>), ASCQ: 0x1, FRU: 0x0
Jul 22 08:37:21 wtn450 scsi: [ID 107833 kern.warning] WARNING: /pci@1f,4000/scsi@3/sd@1,0 (sd1):
Jul 22 08:37:21 wtn450 Error for Command: read(10) Error Level: Retryable
Jul 22 08:37:21 wtn450 scsi: [ID 107833 kern.notice] Requested Block: 1715568 Error Block: 1715719
Jul 22 08:37:21 wtn450 scsi: [ID 107833 kern.notice] Vendor: FUJITSU Serial Number: 0132M62300
Jul 22 08:37:21 wtn450 scsi: [ID 107833 kern.notice] Sense Key: Media Error
Jul 22 08:37:21 wtn450 scsi: [ID 107833 kern.notice] ASC: 0x11 (<vendor unique code 0x11>), ASCQ: 0x1, FRU: 0x0
Jul 22 08:37:21 wtn450 scsi: [ID 107833 kern.warning] WARNING: /pci@1f,4000/scsi@3/sd@1,0 (sd1):
Jul 22 08:37:21 wtn450 Error for Command: read(10) Error Level: Fatal
Jul 22 08:37:21 wtn450 scsi: [ID 107833 kern.notice] Requested Block: 1715568 Error Block: 1715719
Jul 22 08:37:21 wtn450 scsi: [ID 107833 kern.notice] Vendor: FUJITSU Serial Number: 0132M62300
Jul 22 08:37:21 wtn450 scsi: [ID 107833 kern.notice] Sense Key: Media Error
Jul 22 08:37:21 wtn450 scsi: [ID 107833 kern.notice] ASC: 0x11 (<vendor unique code 0x11>), ASCQ: 0x1, FRU: 0x0
Jul 22 08:37:21 wtn450 md_stripe: [ID 641072 kern.warning] WARNING: md: d1: read error on /dev/dsk/c0t1d0s0
Jul 22 08:37:22 wtn450 md_mirror: [ID 104909 kern.warning] WARNING: md: d1: /dev/dsk/c0t1d0s0 needs maintenance

Follows outputs from "eeprom", "format", "metastat -p", "metadb",
"prtvtoc /dev/rdsk/c0t0d0s2", "prtvtoc /dev/rdsk/c0t1d0s2", "df -k" and
"/etc/vfstab"

eeprom: Both disks bootables:
boot-device=disk0 disk1

format: (DISKS INSTALLEDS)
AVAILABLE DISK SELECTIONS:
       0. c0t0d0 <Symbios-StorEDGEA1000-0301 cyl 17344 alt 2 hd 64 sec 64>
          /pci@1f,4000/scsi@3/sd@0,0
       1. c0t1d0 <Symbios-StorEDGEA1000-0301 cyl 17344 alt 2 hd 64 sec 64>
          /pci@1f,4000/scsi@3/sd@1,0

metastat -p (MIRROR STRUCTURE)
d3 -m d0 d1 1
d0 1 1 c0t0d0s0
d1 1 1 c0t1d0s0
d6 -m d4 d5 1
d4 1 1 c0t0d0s1
d5 1 1 c0t1d0s1

metadb (REPLICAS)
        flags first blk block count
     a m p luo 16 1034 /dev/dsk/c0t0d0s5
     a p luo 1050 1034 /dev/dsk/c0t0d0s5
     a p luo 16 1034 /dev/dsk/c0t1d0s5
     a p luo 1050 1034 /dev/dsk/c0t1d0s5

prtvtoc c0t0d0:
       0 2 00 0 66846720 66846719
       1 3 01 66846720 3862528 70709247
       5 0 01 70709248 327680 71036927
prtvtoc c0t1d0:
       0 2 00 0 66846720 66846719
       1 3 01 66846720 3862528 70709247
       2 5 01 0 71041024 71041023
       5 0 01 70709248 327680 71036927

df -k (MOUNT POINTS)
Filesystem kbytes used avail capacity Mounted on
/dev/md/dsk/d3 32910886 17038499 15543279 53% /
/proc 0 0 0 0% /proc
fd 0 0 0 0% /dev/fd
mnttab 0 0 0 0% /etc/mnttab
swap 2242944 24 2242920 1% /var/run
swap 2243752 832 2242920 1% /tmp

/etc/vfstab (MOUNT POINTS)
fd - /dev/fd fd - no -
/proc - /proc proc - no -
/dev/md/dsk/d6 - - swap - no -
/dev/md/dsk/d3 /dev/md/rdsk/d3 / ufs 1 no -
swap - /tmp tmpfs - yes -
_______________________________________________
sunmanagers mailing list
sunmanagers@sunmanagers.org
http://www.sunmanagers.org/mailman/listinfo/sunmanagers

Next message: rich p: "jumpstart cdrom and nfs"

Previous message: Glenn, May (COMFIN, VFS, Consultant): "SUMMARY: Prom Password Lock"
Messages sorted by: [ date ] [ thread ] [ subject ] [ author ] [ attachment ]

Relevant Pages

SUMMARY: metadevice problems
... metacleared all filesystems that were on the external storage array ... recreated the d10 metadevice mirror ... shutdown the server and replaced the two bad disks ... > correspond to the real disk numbers. ...
(SunManagers)
SUMMARY: metadevice naming scheme
... What metadevice naming scheme do you recommend/use? ... with more than one disk controller. ... "...The most recently evolved naming scheme uses mirror names in the d10-d99 ...
(SunManagers)
SUMMARY: Urgent: Mirror Disk Needs Maintenance
... One of sub mirror disks was under state "Needs Maintenance" ... metareplace or metareplace -e as suggested by the metastat was useless. ... The disk was not replaced by a new one. ... One disk of my disk array is under status "Needs Maintenance". ...
(SunManagers)
SUMMARY: D1000 power failure with Disksuite: how to restore to running state?
... disk that the system attempts to read will fail. ... only 1 disk will enter the 'maintenance' mode. ... entire metadevice is taken offline. ... the Disksuite user manual makes reference a power ...
(SunManagers)
Registry File Failure BSOD after RAID repair
... SBS2K with Adaptec 1200A IDE RAID 1 ... After replacing a failed disk, the RAID 1 duplicated the info to the new ... mirror disk from the existing survivor. ...
(microsoft.public.backoffice.smallbiz2000)