media verification issues

[Martin999]

Guys,

Can anyone shed some light on causes of media verification errors.

Running on ...

Solaris and HP servers / storage nodes, >networker 7.1 and IBM LTO drives, 1,2 and 3.

Aware of some causes, eg. scsi probes, tape issue, drive issue but does anyone have any idea how to sort what's what.

We have >450 physical drives (hundreds if you consider DDS), >18000 tapes and around 2000 clients, many being storage nodes. On "average" we're probably getting x100s of media verification errors/week, though out %success rate is high >99%. I'm a bit stuck as to where to start. Sometimes it's the same tapes pooping up again, so mount it up and guess what, no problem. Could perhaps be a slightly "off" drive, so writes ok, but other drives cannot read, which at first glance makes it look like a media issue.

I'm struggling on this one, mainly due to the size of the issue.

thx.

Martin

 

[ag100]

Hi - I would start by taking a few of the tapes and looking for any consistent problems. Are they bombing on you in the same drive(s)? Maybe from the same tape series? Can you relabel one of those tapes and try it in another drive? Anything at all that will let you start to play connect the dots?

I had something like this happen some time ago, and we EVENTUALLY traced it back to 1 HBA that was going bad, that impacted multiple drives.

Good Luck!

-ag100

 

[Martin999]

Will do, thx