Indexing issues - Service wont stop and upgrade info

[James_Slack]

Hi,

Vault 11 on 2008 R2.

Indexing Service

I was having some user issues so I though I would restart my server. However, the Indexing service is refusing to stop and I don't really want to kill it.

It's been about 30 minutes and I am still getting:

"The Indexing Service is pending shutdown. Requesting additional time for shutdown routines "

How do I troubleshoot this?

TYPICAL! This has just completed with a

"The Indexing Service could not complete all the required shutdown routines in the allocated time. Service will be stopped anyway. "

While this is now stopped - anything I can check to ensure it doesn't happen again? Might tie in to below...

Index Upgrades

Also, I am wondering if it has to do with Index upgrades that have been taking place and failing - maybe I was doing something wrong?

I was going in to Manage Indexes

Upgrade

Adding some indexes

Creating the new task and telling it to ignore items that can't be indexed.

Then in the monitor tasks, I have in the past seen mostly seen successful, with a few unsuccessul.

But do I need to do anything to finish these off once that completes?

Also, I scheduled 2 tasks last night of about 100 indexes in each... every single one failed and I can't delete that task.

I am seeing log entries like:

"There are 21 indexing tasks requiring user interaction.

List of tasks (Archive name/Task type):

EV_INDEXING_SUBTASK_TYPE_UPGRADE user name"

Any help would be great.

Thanks

James

[GertjanA · Accepted solution]

Hello James,

I have the same when trying to stop the indexing services. It takes forever. It will shutdown eventually. This might have to do with the (concurrent?) upgrade running.

What I have done to sort of work around this:

Upgrade the archives with the smallest amount of items first. I did batches in which I moved up with 1000 (i.e. 0 to 1000, 1000 to 2000 etc) Eventually, you only have a few large ones left. If you have Journal Mailboxes, configure a new archive and use the new archive before upgrading the old archives indexes.

The fact that it says ' requires user attention" indicates there is an issue with the upgrade, and you need to check inthe Task Monitor page. Click the task, see what state the upgrades are in. You can rightclick a subtask, and select retry, continue etc. I'd continue...

Also verify you run the upgrades OUTSIDE of your archiving window, or stop archiving (if you can) for the time you are upgrading the majority of your users. Out of curiosity, what are the hardware specs of the server?

[James_Slack]

Hardware is good, fairly new 8 core server with 48 gb ram.

Turns out our indexing was toast, called support and they had to rebuild a lot of it, not too sure what they did but they had to delete some stuff in IIS and rename a few folders to get it to rebuild some caches etc. and run some command form the media.

Since then it seems to be working much better.

We always run indexing outside of the archiving schedule, which is a pain as we have so much of both to do - we are struggling to keep exchange afloat with enough space at the moment.

Thanks for the info on the tasks though, that will come in handy now things are working again.