While working on a routine CRM 2011 installation, all system checks passed except for Microsoft SQL Server’s. It failed with “The SQL Server ‘Server Name’ is unavailable. Gut instinct was firewall issues here.

 

 

 

 

 

 

 

 

 

 

 

 

Disabled Windows Firewall on SQL an CRM servers but still received the same error message. Error logs did not provide much to go on here. Ports and connectivity all checked out fine – even added the config database into SQL manually and the system checks were able to determine that the database already existed, throwing the error for that (then deleted it).

Thinking the error may be a bogus one, I set the Ignore Checks key in the registry to bypass the error. This let the installation proceed but ended up failing during the creation of the organization database (e.g XXX_MSCRM). CRM itself installed, created the MSCRM_CONFIG database, and I was able to pull up the Deployment Manager to see the organization state as ‘Failed’. This was the confusing part – how was one database created but the other failed? Rechecked the error log on the installation and was provided with:

Error| Install exception.System.Exception: Action Microsoft.Crm.Tools.Admin.CreateDatabaseAction failed. —> System.IO.IOException: The network path was not found.

After uninstalling CRM and doing a little research on this, it seems similar errors are thrown when drives are not shared appropriately. On the SQL server, I opened a command prompt to run a net share on the drive and was rejected with the “The Server service is not started.” When prompted to start it, system error 1508 was shot back stating that Server service was disabled.

Opened up services, went down to the Server service and did see that it was disabled. Enabled the service and started it, and ran the system checks for the CRM installation again.

Voila! All green. Installation proceeded normally and finished successfully.

Thanks to Gage Pennisi, my fellow Tribridge Co-Worker for spending the hours to determine what the issue was. This issue does not happen on a fresh install of the OS, following best practice. The customer’s servers were built from an “image” that had the service turned off.