NetXMS Support Forum

Hi,

i have found that agents do not restore their connection to the server automatically although i have configured them to use watchdog.
i came across situations that the server crushed for some reason or after a power outage when the server started after the agents and i had to manually restart the agents.

Regards,
Lindeamon

Hi,

connections are from server to agents so it is server supposed to restore connections. Did you try manual status or configuration poll on any node?

Best regards,
Victor

Hi Victor,

yes, i can restore the connection manually but i do not want to do that, i want it to be manually.
although i have 4 agents, i have created a container with filter script to capture all the nodes with agents and if i won't do a manual restore the bindings will be removed.
i guess that this can be considered as a feature request.

Best Regards,
Lindeamon

I'm not sure if this is related, but each time I restart my machine running the NetXMS agent, the server loses connectivity with it permanently until I manually restart the agent (sudo service nxagentd restart).

I am running v1.2.16 on both the server and agent.

More specifically, after I restart the machine running the agent (running Debian 6), the "Last Values" view for this agent's DCIs display "<<ERROR>>" in NetXMS Management Console.

I have tried polling configuration, status and the other polling options, but this does not help.

It seems as if the agent is not starting properly when the Debian 6 OS starts up, but then running "sudo service nxagentd restart" on the agent machine seems to clear things up.

I have looked at the agen'st log file and it indicates that the agent is started OK when the Debian OS machine boots up.

Any ideas? Should I report this as a separate issue, instead of coupling it with the original poster's report here?

Hi,

can you check if TCP port 4700 is listening on this machine after initial agent bootup? If yes, is it reachable from server machine?

Best regards,
Victor