Hyperic 5.0.0 does not see availability change!

[oizik]

Hi All,

We have recently archived our Hyperic 4 installation and switched over to a brand new installation of Hyperic 5!

As a basic test I have setup the agents on a couple of servers and am trying to monitor the availability change by switching the HQ Agent service on and off.

Initially when added the Agentt to the Server the availability has come up as 100% as expected, however now, when I switch the Agent off the HQ Dash still reports that the availability is 100%! I can not seem to figure out how this happens?

Can anyone please shed some light on the matter!

Thank you!

garyj

[admin]

Hello,

Congratulation on upgrading your environment!

Are you using the VC plugin?

Is this the availability you are talking about?

Thanks!

Dan.

[oizik]

Hey Dan,

Basically we have a few servers at out remote sites, all connected to head office with VPNs. What I am basically monitoring for the time being is whether the server is up or down. In Hyperic 4 as soon as the VPN dropped off, I would get an email notification and the Hyperic HQ Dashboard would show 0% availability. However what I am finding now, is that this is not the case. For example about half hour ago I purposely killed a VPN at one of the sites, and the HQ Dashboard still shows up 100% availability and not notifications are being sent.

Gary

[admin]

What is the flavour of the problematic platform that we are talking about (the one that availability is incorrect)? Is it a VM?

[oizik]

Hi

It's a mix of windows 2008r2 and 2003 servers!

The hyperic server runs on windows 2008 r2 which is hosted by a vm ware hyperisor

GJ

+61403265135

[admin]

And are those servers also VMs?

[oizik]

No they are Xeon based servers and windows runs directly on the hardware with no virtual machines!

GJ

+61403265135

[oizik]

Further to my email, it seems to me that it could be some sort of a polling issue. As when I bring the agent down the HQ server does eventually report that it has gone done, but not until hours later!

As soon as the agent is started again the server instantly sees that it has started and communication flows!

Any light on the matter would be much appreciated!

Thank you

GJ

[admin]

Do you see any similar behavior with other metrics?

Can you please check if there is time difference between the HQ server and the monitored platform?

[oizik]

Hi Danzada

Sorry for the late reply I have been away for a week!

I have checked the times and the time zones on the machines, and they are the same, I have the syncing of the same NTP servers.

What I’ve noticed is that if I for example stop SQL Server on one of the remote machines Hyperic within a few minute would display that the SQL server is unavailable, however if I stop the entire Agent on the same remote server it will not be for a few hours that Hyperic would actually display the machine as unavailable.

For example yesterday at 19:00 I stopped the Hyperic Agent on 1 of my severs, and only at 2:02 Hyperic Server generate an email to me with the fact that the machine (Hyperic Agent) was unavailable.

Very puzzling I’ve never seen this with Hyperic 4

Any thoughts?

Thank you

GJ