Hi,
I have a box with two instances of cacti running. One is 0.8.7b polling at 5 minutes interval, one is 0.8.7g polling every minute. Both poll with spine.
This is definitely not optimal but I have to work with it
The issue is that graphs using the exact same setup (i.e. data input, data source, data template, graph template) are not showing up the same between the two versions. The data being polled is in fact a remote custom script execution through SSH.
Script runs fin from cli with the cacti user, output is a simple counter. For some reason, value seems to be anywhere between twice and three times what it should be when looking at the 5 minutes cacti, and value is ok in the 1 minute cacti.
The issue seems to be coming from the polling process:
1 minute (0.8.7g)
09/05/2011 10:49:04 AM - SPINE: Poller[0] Host[33] TH[1] DS[697] SCRIPT: /usr/bin/ssh admin@184.12.5.11 -p22 '/var/emhome/admin/slinksmon.sh', output: 18
5 mintes (0.8.7b)
09/05/2011 10:55:13 AM - SPINE: Poller[0] Host[326] DS[6907] SCRIPT: ssh admin@184.12.5.11 -p22 '~/slinksmon.sh', output: 64
Obvious answer is that one command has complete path and the other does not, but I changed that to check if behaviour in the 5 minutes was better with or without, and change does not alleviate the problem.
Anyone have any ideas? I'm a bit lost here..
Cheers !
[SOLVED]Spine custom script issue
Moderators: Developers, Moderators
-
- Posts: 4
- Joined: Wed Oct 13, 2010 10:07 am
[SOLVED]Spine custom script issue
Last edited by shadowwisps on Wed Sep 07, 2011 5:24 am, edited 1 time in total.
-
- Posts: 4
- Joined: Wed Oct 13, 2010 10:07 am
Re: Spine custom script issue
Just to confirm the last point in my initial post, 5minute/0.8.7b log :
09/05/2011 11:10:12 AM - SPINE: Poller[0] Host[326] DS[6908] SCRIPT: /usr/bin/ssh admin@184.12.5.11 -p22 '/var/emhome/admin/slinksmon.sh', output: 62
Still wrong, even with full paths
09/05/2011 11:10:12 AM - SPINE: Poller[0] Host[326] DS[6908] SCRIPT: /usr/bin/ssh admin@184.12.5.11 -p22 '/var/emhome/admin/slinksmon.sh', output: 62
Still wrong, even with full paths
-
- Posts: 4
- Joined: Wed Oct 13, 2010 10:07 am
Re: Spine custom script issue
Bump, still stuck with this :'(
-
- Posts: 4
- Joined: Wed Oct 13, 2010 10:07 am
Re: Spine custom script issue
Hi
Solved this one: the added sessions are in fact legitimate. The cacti polling every 5 minutes generates about 30 sessions on my host. Much less information is polled by the 1 minute cacti, so very few sessions are generated.
I double checked this by:
1- Creating a perl script to make sure the returned information was parsed ok
2- Running by hand the session poll script on the remote host just at the same time as the 5 minute cacti did it's polling.
Solved this one: the added sessions are in fact legitimate. The cacti polling every 5 minutes generates about 30 sessions on my host. Much less information is polled by the 1 minute cacti, so very few sessions are generated.
I double checked this by:
1- Creating a perl script to make sure the returned information was parsed ok
2- Running by hand the session poll script on the remote host just at the same time as the 5 minute cacti did it's polling.
Who is online
Users browsing this forum: No registered users and 4 guests