02-14-2008 08:35 AM
User tracking has stopped working and I see the following logs in /var/adm/CSCOpx/log/ut.log
Anyone what could be causing these connections refusal?
messages will remian logged to file: /var/adm/CSCOpx/log/ut.log
2008/01/26 10:00:00 main ani MESSAGE ProcessInitializer: Properties will be read from /opt/CSCOpx/campus/etc/cwsi/ANIServer.properties
I= 0value *.*.*.*
I= 1value 6
I= 2value 1
2008/01/26 10:00:03 main ani MESSAGE DBConnection: Created new Database connection [hashCode = 7332084]
2008/01/26 10:01:33 DBConnecton-Reaper ani MESSAGE DBConnection: Closed Database connection [hashCode = 7332084]
Couldn't get I/O for the connection to: hostnamejava.net.SocketException: Connection reset by peer: Connection reset by peer
Couldn't get I/O for the connection to: hostnamejava.net.ConnectException: Connection refused
UTM is shutting down
02-14-2008 08:45 AM
Assuming "hostname" is the correct hostname of the server, then check to see if ANIServer is running.
02-14-2008 09:36 AM
02-14-2008 09:48 AM
Is the hostname referred to in the log correct for this server? UTM tries to connect to that hostname on TCP port 14004 (which should be occupied by ANIServer). Since that connection is being refused, and ANIServer is running I assume the hostname is wrong.
Other causes could be a bad ANIServer.properties, or an ANIServer init problem. If the hostname checks out, please post NMSROOT/campus/etc/cwsi/ANIServer.properties and the ani.log.
02-14-2008 11:59 AM
02-14-2008 12:24 PM
Please post the ani.log and the output of pfiles on the ANIServer PID.
02-14-2008 12:41 PM
Joe
From looking at the ani.log is looks like it has started working at 2PM after the third time I stopped/started the daemon. What could be wrong with the db just to stop working? Can I configure some types alerts that send the notification when certain services die?
I'm on Sol5.9, CS3.0.5, RME3.0.5.
Thank for all your help and time.
02-14-2008 01:03 PM
The problem was not with the database, but with ANIServer. There isn't any built-in way to get notified when a process goes down. You'd have to script something that looks at pdshow output. That said, some processes may legitimately go down in normal operation (e.g. UTMajorAcquisition, DeviceDiscovery, DataPurge, etc.). So you might end up generating a lot of false positives.
From your pdshow, it looks like the daemons were restarted today at around 11:56 local time. It can take up to 30 minutes for everything to properly initialize. the problem may have just been transient.
Discover and save your favorite ideas. Come back to expert answers, step-by-step guides, recent topics, and more.
New here? Get started with these tips. How to use Community New member guide