I wrote about this new feature we added to control this in SCOM 2007 R2 with the advent of CU4: http://blogs.technet.com/b/kevinholman/archive/2011/02/07/a-new-feature-in-r2-cu4-reconnecting-to-sql-server-after-a-sql-outage.aspx
A scenario might occur where a SQL outage exists, and then the management servers will not automatically connect to SQL again once it comes back up on the network. If you have experienced this you should consider applying this resolution.
However – those registry locations have changed in OM 2012. This was done to sync the Data Access layer between Service Manager and SCOM. There has been some confusion if these configuration settings are supported and/or work in SCOM 2012.
The new locations are documented at: http://support.microsoft.com/kb/2913046/en-us
Start the Registry Editor.
Locate and then click the following registry subkey:
Create the following two new DWORD values:
Decimal value: 1
Decimal Value: 60
DALInitiateClearPool should be set to Decimal value “1” to enable it.
DALInitiateClearPoolSeconds should be set to Decimal value “60” to represent 60 second retry interval. The DALInitiateClearPoolSeconds setting controls when the management server drops the current connection pool and when the management server tries to reestablish an SQL connection. We recommend that you set this setting to 60 seconds or more to avoid performance issues.
Restart the System Center Data Access Service, and the System Center Management/Microsoft Monitoring Agent service on the management server for these settings to take effect.
This solution applies to SCOM 2012, 2012 SP1, and 2012 R2. Along with the same versions of Service Manager 2012.