Replaced bad disk now getting last erred on another HELP!

[philpe]

Hello,

I have an old Sun Enterprize 1 server configured with a Raid 5 device using disksuite. Just today, I had a disk go in to the maintenance state. The format command couldn't even tell what kind of disk it was, so we assumed the disk was bad. We shut down the server and replaced the bad disk. When we brought the system back up, the format command now shows the right disk type. I fmthard the disk to make it like the others in the raid group, then use the metareplace -e command to enable the drive again. metastat shows everything as okay, except the one I replaced which in doing a resync. Then a few minutes later, the resync stops, and another drive in the raid set has a status of "Last erred".

Now everywhere I read, says I need to metareplace the disk in maintenance state first, but it just doesn't what to go, it always just stops the resync after about a minute. The messages log lists a read error on the drive in "last erred" state. What do I need to do to get the other disk resynced? The data is still available, so I've tared it off to another system for now. Any ideas???

Thanks

 

[Solfan]

In a raid 5 stripe you can only have 1 bad disk/stripe. It appears that in this case you have two. The first caused the stripe to go into "Needs Maintenence" mode the 2nd resulted in a "Last Erred" mode. Therefore this raid 5 stripe is unusable resulting in the problems that you are having.
So unfortunately it appears that you will need to:
1. Replace the bad disk(s)
2. Set them up some disksuite can used them.
3. Restore data from your backup media.

Hope this helps!

 

[philpe]

Thanks Solfan,

I seemed to have worked around the problem, The way disksuite works is if a second disk goes bad in a raid 5 group, it makes it as you said Last Erred, but still allows you to read from the disk. I read somewhere that if you do a read analysis of the disk you may be able to fix some errors. I did just that, and it repaired about 4 problems on the disk. Then I was able to resync the new disk. It also says that you may already have some courrpt data because of the Last Erred problem, but that's better then nothing. Not sure what caused the last erred problem, could have been there for a long time, but only showed up when I tried to do the resync.

I hope no one else ever has to deal with this kind of problem, but my suggestion would be to make a backup of your data as soon as you get a single bad disk in your raid 5 group. Could save you alot of trouble when you go to replace it. Thanks again for your help.

Phil