I recently experienced the exact same problem. The polling server lost connectivity with the SQL server and the front page came back up but I didn't realize that the polling had not resumed until the next morning. I wasn't sure if this was an isolated incident or not (obviously not). It doesn't seem like this used to happen in the past when connectivity was lost. Is this a new bug or has this always been the case?
I have seen this same issue. We have our Orion Database on a SQL cluster. Sometimes during a cluster failover Orion would work but it wouldn't record any data. I have adjusted the database timeout to help prevent this during a cluster failover.
I noticed the same problem but really the horrible part is all of the alerts go off so I get several thousand messages all at once. My BB can't keep up.
We've seen the same thing. There was a problem with our SQL server where the main process would hang and our DBA's ended up killing it. Our server admin noticed an error on our Orion server and when it went away everything was happy. I checked the the polling engine stats and it had been over 17 hours since it talked to the database but everything seemed fine.
I ended up killing then restarting the NetPerfMon service and it came back. (Note: I always have to do an end process on it, it NEVER shuts down on its own.)