We started having a problem on 3/19 - our production server was halting.
Initially this appeared in the Event log as a series of cascading MSMQ failures caused by a login failure from a remote server. The strange thing about this login failure what the login name, it was ''. (Blank).
Further analysis shows the cascade was a symptom of TCP/IP halting, apparently due to the cascading errors raised when '' failed to login.
After eliminating any job making a remote call to the production server from the remote server, I ascertained that the only other thing that could possibly be wanting to login to the production server from the remote server was SQL Response, as the remote server hosts the repository
I halted the service on the remote server yesterday and we ran fine.
This morning I came in and started the service and in short order the production server halted again.
Making this more problematic - the production server is running a software package that is on its last legs, and we are not supported by the vendor. We are starting a migration to a new system next month, but for now, I cannot do much meddling with updates and such to the production server.
I hate to not use SQL Response as it has been invaluable to us here for the past year plus - but I cannot keep having everyone log off to restart this server 2 or 3 times a day.