Server Management Metrics problem

Janne Haatainen's picture
Janne Haatainen
October 28th, 2009

Hi!

I'm evaluating Symantec Server Management Suite and I'm having problems getting information from my servers.

From my ESX-server, I get nothing. Or, I do get information, but it's only 0, when I view historical data, all I get is a straight line and all values are 0. Monitor Policy packs used are "ESX- Basic" and "ESX - Extended Host". I try to get values from CPU, Memory and Disk I/O but all values I get is 0.

And from Windows (w2k3 & w2k8) I get values with "Memory - Available (M)Bytes" and "Memory - Committed Bytes" by using Agent-based policy. With Agentless policy I also get CPU and Network etc values, but with agent-based policy I can't get CPU or Network values..

Any ideas why this is happening?

Regards,
Janne

Filed under: ,
0 votes
MonitorMan's picture
MonitorMan
3 weeks 3 days ago

Janne If you run the shell

Janne

If you run the shell statements within a console session on the ESX server, do they return actual expected values?

Third Floor Garage Door Operator

0 votes
Janne Haatainen's picture
Janne Haatainen
3 weeks 3 days ago

Yes, for example if I run

Yes, for example if I run command wich is in metric "Processor Idle Time% ESX", I get values like:
"16 40 9 36 0.13 16007"

If they are expected values, that I have no idea :) But at least it gives me values other than zero.

0 votes
MonitorMan's picture
MonitorMan
3 weeks 2 days ago

Joseph will probably chime in

Joseph will probably chime in here, but I'll give it a stab. So I'll share what I do with my ESX 3.5 Update 4 clusters to help in troubleshooting monitor solution.

1) Are we certain the servers are getting the MS policies? You can check the metriclog.xml in the \opt\altiris\notification\monitor\var (IIRC) folder for entries and growth to see if the applied policies are being queried.

2) Pretty sure the monitor.log in the same folder can help you diagnose connection issues. You may need to turn up the logging though.

3) In the bin folder of MS, there is rcscript to check the status to ensure the MS daemon is running

4) If you want to get creative, you can leverage aex-metricprovider -c to output what MS is doing to a console session. You can increase MS logging with aex-metricprovider -l. A -h will show you the rest it can do.

Just some ideas to ensure that the MS daemon is actually processing the policies. Obviously I would also check the esx firewall, ns agent communication, filter placement, etc.

Third Floor Garage Door Operator

0 votes