Qfyilyi

I have a SQL Server named instance that runs the standard syspolicy_purge_history job every night. The job is marked as succeeding every night, but every night on the DEFAULT instance on the same server, there is a failed login connection from the service account of the named instance.

I've tracked it down to the Erase Phantom System Health Records step of the syspolicy_purge_history job. The job uses the standard code and does not appear to have been modified in any way:

if ('$(ESCAPE_SQUOTE(INST))' -eq 'MSSQLSERVER') {$a = 'DEFAULT'} ELSE {$a = ''};

(Get-Item SQLSERVER:SQLPolicy$(ESCAPE_NONE(SRVR))$a).EraseSystemHealthPhantomRecords()

There are two other named instances on the same server, running the same job with the same code, and they do not generate failed login errors on the default instance.

I tried deleting the job, and recreating it by scripting the job from one of the other named instances. It made no difference.

I tried changing the code of step 3 to:

(Get-Item SQLSERVER:SQLPolicy[MyServerName][NamedInstanceName]).EraseSystemHealthPhantomRecords()

and alternatively to:

(Get-Item SQLSERVER:SQLPolicy[MyServerName][NamedInstanceName]`,[NamedInstancePort]).EraseSystemHealthPhantomRecords()

Both versions still generate the failed login on the default instance.

Both versions of the command appear to work successfully if I run them from a powershell window, and do NOT generate a failed login.

I cannot think of anything else to try. Google seems to be out of suggestions for me too.

Any ideas?