This post is more than 5 years old
2 Posts
0
615
November 1st, 2009 19:00
Indexes Showing Blue Exclamation Marks in EmailXtender Administrator
I have been reindexing by manual methods on our Xtender 4.81.409
to resolve these issues but this seems to be a never ending problem.. with constant message corruptions or indexing faults?
there are several trains of thought in regards to using healthchecker as this causes known issues .. so i have been using the manual corrupt.idx
is there any general advice or info that can be linked ?
i am reading through our admin guide for this product ..
to resolve these issues but this seems to be a never ending problem.. with constant message corruptions or indexing faults?
there are several trains of thought in regards to using healthchecker as this causes known issues .. so i have been using the manual corrupt.idx
is there any general advice or info that can be linked ?
i am reading through our admin guide for this product ..
No Events found!
RKatwal
2 Intern
•
600 Posts
0
November 3rd, 2009 21:00
There are many reasons why indexes can get corrupt. Here are few of them:
1. Write Delay:Some times due to any reason a thread is unable to write to the disk and index thread(exIndex.exe) terminates after some time. In such cases when healthcheck will run it will report the issue because number of messages archived will be more than the number of messages indexed. If you are trying to do reindexing and data is purged using DX on another media then it could be adding while reindexing data because of the delay in fetching messages if direct read is enabled.
2. Unexpected server shutdown: Some times if someone does an exexpected shutdown or server restarts unexpectedly, you can see probelm for index sets as ExIndex.exe might have died while updating index set.
3. Disks locked while backup: It is expected that EX should only be backed up after running ExSuspend.exe command line. This utility allows you you to have error levels returns in case it cannot suspend something and also it allows you to set wait interval.If backup software locks down disk or files that can cause problem for index threads as I mentioned earlier. I do not think EX is VSS aware so try not using that to backup EX.
4. Anti Virus not configured properly: Even though when EX would have been installed, it would have been taken care to make sure that AV do not scan any of the EX directories but with the time Admins tend to move directory structure or make changed to the AV scanning policies organisation wide and forget about EX server. If AV is scanning EX folders that will cause problem for indexes. Some times you run into this situation after a particular update and while doing reindexing because in past the same attachment might not have qualidied as a problem but with updated signatures for AV, now the same file could be getting removed by AV when it was getting extracted to be indexed.
5. Healthcheck marking index sets with blue ! marks because it ignored open voluymes: There was a engineering escalation open in past where healthcheck was not counting open volumes and it was marking open volumes with blue mark. Open volumes can have more messages archived and less messages indexed because SQL and archive would add data to volume streight away and indexing happens after that. Latest version of healthcheck do not have problem. I am not sure but I believe 4.81.409 was affected with this issue.
These are few out of many user cases. I will also suggest you to move to 4.81.1147 at least because there were fixes that were related to indexing and healthcheck in last few patch kits. If you will go throgh all the fixed defects for 4..81.1147 you can see all the fixes that were added in the releases between the one you are running and the latest one.
There is a support note on common causes for indexes with missing messages you should be able to search for the support note on powerlink.
I hope that helps.
Thanks,
Rajan