Disk graph gap / incorrect value

dorje · Post by **dorje** » Wed Dec 01, 2004 9:30 pm

Team,

Problem
Stocastic graph with gaps and values that sometimes are showing used amount of space / sometimes total disk size - even though should only be graphing used amount.

I've looked at the forum for anything similar but have not found anything that I think relates.

Setup
- cacti 0.8.6b on Fedora Core 2 (FC2)
Polling a win2k machine for disk space using
- net-snmp 5.x / version 1
- cactiD / 6 processes / 12 threads
Host template = Windows2000/XP

PS Many thanks to one and all for Cacti - great tool - great community.

_JPL_ · Post by **_JPL_** » Thu Dec 02, 2004 2:45 pm

I've been having the gapping out problem ever since upgrading to 0.8.6b on FC2. Still no solution to the gapping problem. You're not alone.

dorje · Post by **dorje** » Thu Dec 02, 2004 3:21 pm

_JPL_ many thanks - I've seen your other posts but thought that my situation was a little different (though probably not).

For clarity I've posted (below) all the graphs for this host.
As you'll be able to see there are only three graph types exhibiting this gap / high / low stoccastic behaviour - Disk Space (which is not right), cpu utilization (which I kind of expect) and Ping latency (which is also not right - the cacti server and the host shown in the graphs are on the same subnet).

If I hear / work out anything I'll let you know.

_JPL_ · Post by **_JPL_** » Thu Dec 02, 2004 3:36 pm

Yeah, now that I see the other graphs it doesn't appear to be the same problem, but that first graph sure rang a bell with me.

I too, will let you know if I fix my gap problem. While it may not be the same it could still be useful.

mrh · Post by **mrh** » Tue Mar 08, 2005 10:45 am

Hi guys,
I've the same problem when monitoring certain win2000 hosts with the disk usage template (under win2000/xp host template). I've noticed that when I remove the graph not drawing propperly then the next one begins to draw incorrect results. If I remove this again and back to create, then it will draw correctly but not the next disk graph... very weird!!! (cascade problem)

I will thank very much any help or comments.

MRH

Post by **rony** » Tue Mar 08, 2005 11:30 am

Increase your snmp timeout for the affected hosts.

When quering Win2K and WinXP, the disk information causes the agent on the servers to check all the drives. This can lead to timeout problems.

mrh · Post by **mrh** » Tue Mar 08, 2005 12:09 pm

thanks Tony, I've increased it to 5000 and it's still not drawing...

MRH

Cacti

Disk graph gap / incorrect value

Disk graph gap / incorrect value

Full Graphs for this host

cacti gaps problem win32 disk template

Who is online