Hello Users in our Exchange 2007 environment reported that Outlook search, when in online mode, doesn't work properly. I ran the Test-ExchangeSearch command which reported that a bunch of DB's on our server had corrupt search indexes. I understand we need to run the ResetSearchIndex script to fix this.....we have about 5 DB's on the server which need this run against them, is there any issue with i. Doing these at the same time (i.e. running the command for each DB one after the other) ii. Running these during business hours

Before going to do this, you must have full backup and do it in non business hours to face any surprise. Reseed the Search Catalog: http://technet.microsoft.com/en-us/library/ee633475.aspx Anil

This is for Exchange 2007 by the way, not 2010. Why do you say we should run out of hours, what's the possible issue?

If your catalogs are already corrupt then it doesn't really matter whether it is done in off-hours or not. However the crawling can put quite a load on the server which might impact mailbox access, i.e. slow it down. You can run the commands one after another, Exchange will then do the rebuilds sequentially. i.e. It won't try and do them all at the same time which really would slow your server down to a crawl. However if memory serves the script stops the search index service, deletes the CI then restarts the search index service. If doing multiples you'd really be better stopping the service, deleting the catalog folders manually then restarting the service. It amounts to the same thing but means the service is only down once and Exchange won't be part way through a re-index each time you run it. Neill

Hello Neil Thanks for answering. I guess if we do multiple DB's then that is more load on the server isn't it? Do you know what the actual resource hit is (I/O, CPU etc?). So we're probably better of down one, waiting for the crawling to finish then doing another etc? Thanks again

It really depends on the size of your mailboxes and number of items within them. It does each database sequentially as I said so you can purge all the catalogs at once and just let it run if you want. One place I worked we had absolutely massive mailboxes with huge item counts so the crawl would take about 10 hours per database. Not great. It didn't add much to I/O, CPU in itself but we had other problems (those large numbers of items again) so we tended to do one database per night if required. Neill

These servers are actually part of a CCR cluster with an SCR target in another AD site. Does anyone know how this corruption is affected for passive copies of the database?

How to Rebuild the Full-Text Index Catalog http://technet.microsoft.com/en-us/library/aa995966(EXCHG.80).aspx Thanks, Simon