I set up a event action plan to notify via email when a set of servers go online or offline. Now everyday I get spammed by IBM Director with topology online/offline messages when the servers are up and running fine. Has anyone ever seeen this problem
Never seen it before...it may be that although the server is up and running fine, the Director Agent installed on the server is not, or the network connection between the Director Server and Agents has issues. What version of Director are you running? How do the Agents connect to the server? Same subnet? If not, are Relay Agents configured?
I'm running 4.12. The monitored servers are on a separate subnets from the management server. And these servers are the relay agents for thier respective subnets
Are the network adapters correctly configured with a gateway address to reach the Director Server's subnet? What is the network timeout setting (TWGIPCCF.EXE)? If it's the default (15 seconds), try increasing it to 30-45 seconds (also increase it on the server and restart Director in both cases (net stop/net start twgipc).
To confirm whether or not the Director Support Program (twgipc) is really stopping, you can configure a Resource Monitor for the executable.
Thanks for the info it worked great but I've started to monitor some SNMP devices and I've been getting the same problem, Director says the device is offline when it is not. Is there any other way to monitor a network device without using the topology online filter?
The method for discovery of SNMP devices in Director parallels the method for discovery of native Director Agents. It is NOT a “ping spray” of IP devices in the local subnet to find agents listening on port 161. Instead, it uses the IP addresses configured in the Discovery Preferences for SNMP devices as seed addresses and uses the values in the ipNetToMediaNetAddress column of the ipNetToMediaTable variable of the seed addresses’ MIB as the candidate devices to discover.
The default seed address is the local Director Server host system. However, the best practice is to configure a more practical device as the seed address such as a router or a heavily used server such as an http server that is likely to have a lot of values in its ipNetToMediaTable variable.
If the Director Server is newly installed, it’s likely that there will be few addresses in its ipNetToMediaTable variable, and hence selecting Discovery->SNMP Devices will discover few devices out of the box.
To see the values of the ipNetToMediaTable variable, drag and drop the SNMP Browser task on the Director Server host and drill down to iso.org.dod.internet.mgmt.mib-2.ip.ipNetToMediaTable.ipNetToMediaEntry.ipNetToMediaNetAdd
ress.
In this case, either configure a more practical seed address, or perform a discovery of Director Agents which will broadcast to the local subnet. IP addresses that it reaches will be added to the IpNetToMediaTable variable.
Sorry, probably more info than you wanted. To summarize, you'll need Relay Agents for the SNMP subnets too. That should resolve your issue.
This site uses cookies to help personalise content, tailor your experience and to keep you logged in if you register.
By continuing to use this site, you are consenting to our use of cookies.