This is a discussion on stale root partiition - AIX 5.3 within the AIX Operating System forums, part of the Unix Operating Systems category; --> Just migrated a system from 5.1 to 5.3 which went just fine until I decided to mirror rootvg. The ...
| |||||||
| FAQ | Members List | Calendar | Search | Today's Posts | Mark Forums Read |
| ||||
| Just migrated a system from 5.1 to 5.3 which went just fine until I decided to mirror rootvg. The system is at ML2. The mirroring seemed to go fine: 0516-622 : Warning, cannot write lv control block data. 0516-1124 mirrorvg: Quorum requirement turned off, reboot system for this to take effect for rootvg. 0516-1126 mirrorvg: rootvg successfully mirrored, user should perform bosboot of system to initialize boot records. Then, user must modify bootlist to include: hdisk0 hdisk1. But wonder what that first line means... I saw that when unmirroring it to do the upgrade then when I did the unmirrorvg: 0516-622 rmlvcopy: Warning, cannot write lv control block data. 0516-1135 unmirrorvg: The unmirror of the volume group failed. The volume group is still partially or fully mirrored. But when I ran the unmirror again it went fine: 0516-1132 unmirrorvg: Quorum requirement turned on, reboot system for this to take effect for rootvg. 0516-1144 unmirrorvg: rootvg successfully unmirrored, user should perform bosboot of system to reinitialize boot records. Then, user must modify bootlist to just include: hdisk0. So I thought I better check it before I reboot. I did an 'lsvg -l rootvg' and found that all of the logical volumes in rootvg were open/stale. I then ran a 'syncvg -l rootvg' and got a series of errors saying it couldn't syncronize each logical volume. But when I looked again, all of the logical volumes were synchronized except / (root). I tried 'syncvg -l hd4' but it said it couldn't synchronize and hd4 still shows open/stale. I can run rmlvcopy on hd4 fine. I can also run mklvcopy of hd4 onto hdisk1 and it works but still shows as open/stale. When I 'lspv hdisk0' there are no stale partitions but if I run 'lspv hdisk1' it shows 88 stale partitions - which is the amount of partitions for /. Everytime I run 'syncvg -l hd4' I also get DISK_ERR1 errors which IBM says mean the disk is definitely going bad. But I ran a mksysb and run TSM on the system all night and no disk errors popped up at all. I removed the hd4 copy on hdisk1 so it would show as open/sync in case it gets rebooted. Problem is I can't do much with the root partition in a stale state. I can't replace the disk and remirror it from hdisk1 because the root mirror is apparently stale. I am wondering if this disk is really going bad or is it something else since I only get errors when trying to mirror hd4. Am I screwed here or is there some way I can try to fix the root partition so it will mirror? I saw synclvodm but not sure if that would help. Any help is appreciated. |
| |||
| rstorch1@netzero.net wrote: > Just migrated a system from 5.1 to 5.3 which went just fine until I > decided to mirror rootvg. The system is at ML2. The mirroring seemed .... Go for ML03 and the latest CSP see http://www14.software.ibm.com/webapp...e=53&b=5300-03 ( Pretty much APARS about mirror problems ) So apply the 5300-03-ML and 5300-03-CSP Then try to remirror /dev/hd4 If it fails: - Remove mirroring so that al stuff should be left on hdisk0 - Boot from from cd/dvd/ into maintenance mode and try to access hd4 and/or fix fs corruption. All without warranty Hajo |
| ||||
| Le Thu, 06 Apr 2006 07:23:42 -0700, Hajo Ehlers a écrit*: > > rstorch1@netzero.net wrote: >> Just migrated a system from 5.1 to 5.3 which went just fine until I >> decided to mirror rootvg. The system is at ML2. The mirroring seemed .... > > Go for ML03 and the latest CSP > see > http://www14.software.ibm.com/webapp...e=53&b=5300-03 > ( Pretty much APARS about mirror problems ) > > So apply the 5300-03-ML and 5300-03-CSP > > Then try to remirror /dev/hd4 > If it fails: > - Remove mirroring so that al stuff should be left on hdisk0 > - Boot from from cd/dvd/ into maintenance mode and try to access hd4 > and/or fix fs corruption. > > All without warranty > Hajo I second this recommandation... We have had some curious problem about mirroring on ML2 (I have made a lot of manipulation with IBM because the swap partition was not fully mirrored and the funny part was that AIX saw 3 PV instead of 2) Seems a lot is fixed in this ML...I have had some trouble with NIM also with ML2 as well Regards, JYL |