03-07-2011 03:12 AM
Hello ,,
EV 9 installed on Windows 2008 R2 64
Event:The index volume has been marked as failed ..
Event: Unable to repair because of " Too many consecutive failed items " ..
Found a proposed solution in the forum about increasing the registry value HKEY_LOCAL_MACHINE\SOFTWARE\KVS\Enterprise Vault\Indexing\PoisonPillCount
DWORD
Default: 1
But there are no key for Indexing\...... in registry for HKLM\SoftWare\Wow64node\KVS\EnterpriseVault!
Do I add all keys manually ?
regards
/Bo
Solved! Go to Solution.
03-07-2011 05:21 AM
Yes, if the key doesn't exist you need to add it manually. Before you set it though you might want to dtrace and try to find what it is failing on and why if you have not already.
03-07-2011 05:21 AM
Yes, if the key doesn't exist you need to add it manually. Before you set it though you might want to dtrace and try to find what it is failing on and why if you have not already.
03-07-2011 06:46 AM
Agreed with tony, the poison pill count is a cheap hack, if you genuinely have that many missing items, you need to find out genuinely how many are missing, why they're missing, how to get them back or how to clean them up
If you're on EV8 then you need to run EVSVR, if you're on EV2007 or below then you need to run VSVerify
03-07-2011 07:14 AM
PoisonPill setting is a last resort to get the index rebuilt skipping the failed items so you have some access to an index. It is not a solution.
Follow the advice from Tony and Alex and identify the root cause of the issue. If you do a rebuild or repair and DTrace the results then post them here and we can offer advice to resolve the issue
If this is critical and you need access to the index ASAP then open a support case with support.
03-07-2011 03:11 PM
Ok .. here comes a DTRACE LOG(Update Index)
All other indexes works fine ..
/Bo
03-07-2011 03:14 PM
We are on EV9 ..
/Bo
03-07-2011 06:24 PM
you need to dtrace storagecrawler
03-11-2011 05:03 AM
Bo, I'm not sure if it could help, but we could solve some issues regarding failed index rebuilds.
We found the affected index volumes, moved the files to another location, rebuild the index and it works.
In some cases it need some time. ;)
Good Luck!