polling engine problem

Hi All;

help me.. i have NPM v10.1 that running on Windows Server 2003 and my DB is Microsoft SQL 2008 Enterprise Edition that running on Windows Server 2008. My problem is on my polling engine. The NPM always show some nodes are up and after several minutes its will going down and after a several minutes, its will up again. This problem is repeated with stopping. When I ping the nodes IP from that server, there is no network intermittent issues, everything is ok. I have no idea what are the root causes of this problem actually.. Hope somebody can help me.. TQ

Find more posts tagged with

orion_10_npm

orion_10.1_npm

npm_10_polling

npm_10_poller_manage

npm_10_interface_polling

Accepted answers

All comments

jusmax

Good Day,

Have a look at your timing between your SQL Server, Main Polling Engine and the Polling Engine, they all should be within 1 minute of each other. If they are more that 5 minutes apart in timing you get all sorts of strange issues.

Mohd_Rozaini

hi jusmax;

thanks for the reply.. i already confirm the time between both server sql & polling engine are same.. but the NPM still show up and down ..

jusmax

Ok, I guess you have checked these already, however just to have an idea:

1. During the outages, can you successfully ping the end the devices from the said Polling Engine?

2. Are the devices being monitored in the same network as the polling engine?

3. What is the difference in the devices staying up and the devices going down?

4. Were these migrated recently from a previous/main polling engine (check to see if the versions of NPM are the same, how was the migration done if any)?

5. What is the duration of the outage, how often and when did it start happening (if this is not a new installation)

6. Can you see if there is a duplex mismatch between the Server and connecting Router/Switch?

7. If there is a router, are there any ACLs, QoS settings that can block limit SNMP, ICMP.

8. How is bandwidth to the network being monitored? Check for over utilization and broadcasts.

I know you may have already done some of these checks, just making sure all of the bases are covered. Now number 7 above is an interesting one. When the devices go down, do the Interfaces within the devices also drop go unknown or do they stay up? Reason I am asking is that even though NPM uses SNMP for polling and gathering data, it relies on ICMP for Availability and connectivity. If the interfaces are staying up and green when the node shows as down, then you have an ICMP problem getting to that network being polled. Hope this helps.