May I check if anyone encountered spooler on remote monitored server that is not passing alerts to Hub.We had to restart the spooler in order to trigger alarms.
Spooler was running and no error messages during that time
Windows, Unix, Linux or iSeries?
Was there anything in the spooler log?
it happened for unix machine. no error message in spooler logs. It is quite dangerous, as we are thinking server is being monitored, however all alarm messages are stucked.
Ouch! is it the same machine all the time or do you see the same problem on a few robots?
Are you running the robot as root or with reduced rights? It sounds like the sort of thing permissions problems can cause.
Did you upgrade to the latest version? The readme says lots of stability fixes.
I have seen the same issue and it is usally caused by the spooler queue getting very large. I think this can happen when one of the queue files gets corrupted. Of course, if the spooler is unable to send messages then there is no way for it to send you an alarm. I ended up writing a script which reaches out to each robot, checks the spooler queue size, and sends an alarm if it is over some threshold.
Here is the function we use to check the queue. I might have actually gotten this from the forum back in the day.
function SpoolerSize (address) local spoolerinfo, spooler_stat = nimbus.request(address .. "/spooler", "info", nil, 1) local queued = -1 local stat = 777
if spooler_stat == 0 and spoolerinfo ~= nil then queued = spoolerinfo.queued end
Retrieving data ...