False positive in one server, whereas not in the rest

[banderas20]

Hi everyone,

I have several servers based on the same template, as they have the same processes and services running.

In the template I have configured an item as follows:

proc.num[name.exe]

and its trigger as:

{Server1roc.num[name.exe].last(0)}<1

The problem is that the trigger arises, although the process is actually running. So it's a false positive.

I have other servers with the same process and templates, and the trigger works correctly. That is, it doesn't arise.

¿what may be going on?

The operative system is Windows Server 2008 x64.

Thanks in advance.

[kevind]

I see the same problem for processes on CentOS 5.4, in that case it's due to a kernel bug (race condition) that was fixed in CentOS 5.5 and later. I haven't seen this issue for Windows Server 2008 x64 (does it have all updates applied?).


You can filter out transient errors like this by requiring, say, 3 successive errors to fire a trigger (which will delay the alert but you're less likely to get false alerts):

{Server1roc.num[name.exe].max(#3)}=0