cancel
Showing results for 
Search instead for 
Did you mean: 

Journal indexes showing "failed" using IndexVolumeReplay

Kathy_Dean
Level 4
I have 4 journal indexes that all showed "failed" using IndexVolume Replay. 
We also received errors in Discover Accelerator when searching these indexes.
I checked the Avtrace.log files assocated with these indexes and I don't see anything in there about corruption.
 
Our indexes are about 18GB in size and are located on a NetApps device.  I found they are getting backed up but also found that the EV services were not getting shut down before the backup kicked off.
So I don't know how good our backups are???
 
At this point I don't know if I should try performing an "update" on these indexes first.
I am uncertain if I should do a rebuild without knowing how long it will take or the impact it will have on the server.
 
If anyone could provide some insight I would appreciate it.
 
Thanks.
 
Kathy
5 REPLIES 5

TonySterling
Moderator
Moderator
Partner    VIP    Accredited Certified
it would be worth putting a back up copy in place and then do an update first to see if that corrects the problem.

Kathy_Dean
Level 4
Can you tell me how long an update takes?

Rob_Hanson
Level 6
Employee
A lot less than a complete rebuild but hard to say exactly how long.  It depends on the amount of RAM on the box, how fast the processor and how long since the backup.
 
SHould not be more than a few minutes per Archive.
 
Journal Indexes tend to take longer.

Kathy_Dean
Level 4
I choose to run an update  on a journal index that is 6gb in size and it has been running for over 13 Hours?

Michael_Bilsbor
Level 6
Accredited
Enterprise Vault will mark an index as 'failed' if it had a repeated problem opening/ inserting items.  It marks it as failed so that it doesn't repeated log lots of errors when some kind of 'hard' error occurs etc.
 
So when you do an 'update' it basically says to indexing, set the index status back to normal and carry on updating from where it was upto.  So for example say an intermittent error occured then an update would restart the update process and complete.  So it could 10 items need to be added, it could be 20,000 all really depends upon when the index got marked as 'failed'
If a hard error occurs (say a physcially corrupt index) then the update obviously won't complete. However because we try and fail again, errors will be logged and so that in itself is useful to help try and determine the root cause.
 
It is always a good idea to try Update first, as it's quite possible there is no 'hard' issue and it will succeed.
However you should look at errors in the event log to determine why the index got marked as failed in the first place, otherwise the problem may occur in the future again.  Perhaps for example Anti-Virus software was scanning the files etc