I have some empty graphs, when the gatherer runs he makes the graph but without values in, this apens for all my disks on windows and solaris/linux OS, but other graphs like network are working. There are others like system processes, cpu percent system etc that only collect values of my solaris box. I have net-snmp installed on my solaris and windows boxes so this should be running.. do i have to set any parameter for the disks i whant to monitor ? when i recache disks he gives me a list of devices on my system like /var or /proc etc (on solaris) and c, d etc (on windows) i allready added them to monitor but they are all empty no values in.. is there any plugin, patch or something i forget ? When i run the gatherer with -am, there are some erros like
[code144b79c7d4f][Dev: 0005] [Sub: 0042] [Mon: 0111] Starting Monitor.
[Dev: 0002] [Sub: 0026] [Mon: 0078] Value: No Such Instance currently exists at this OID
[Dev: 0002] [Sub: 0026] [Mon: 0078] Value is not numeric.
[Dev: 0002] [Sub: 0026] [Mon: 0079] Starting Monitor.
[Dev: 0005] [Sub: 0042] [Mon: 0111] Value: U
If anyone could help.., tks in advance ?
For Linux and other Unicies make sure that your community you're polling as can access the part of the SNMP tree that contains the processor utilization values. We usually modify the default snmpd.conf to have a line like this
[code17502a50dd7]view systemview included .1[/code17502a50dd7]
Windows boxes use a different OID to gather CPU stats - the default Linux templates won't work for a Windows box.
Disks - right.
For disks you need to modify your snmpd.conf file to include which mounts you would like to monitor; after that the disk cache will show your disks. Add lines similar to this for each of your mounts you would like to monitor
Disks are working, windows boxes ill see later, now i'v added a custom script (check_http -H %ip%) to monitor some http sites, moved it to netmrgs /libexec, and when i run the gatherer,
[code1a1f889b094][Dev: 0002] [Sub: 0049] [Mon: 0147] Starting Monitor.
[Dev: 0002] [Sub: 0049] [Mon: 0147] Sending '/usr/local/libexec/netmrg/check_http -H 192.168.40.8' to shell.
[Dev: 0002] [Sub: 0049] [Mon: 0147] Value: HTTP ok: HTTP/1.1 200 OK - 0.010 second response time |time= 0.010
[Dev: 0002] [Sub: 0049] [Mon: 0147] Value is a decimal.
But what i want is that he takes the URL not the IP, i have %ip% but i would like to have something similar to %url% so that it gives ex " Sending '/usr/local/libexec/netmrg/check_http -H www.nokia.com'
to shell. ", and then graph it whit value 1 if Value HTTP ok and value 0 if there's no response.
You probably want to just use [id15a1d212c]%parameters%[/id15a1d212c] instead of [id15a1d212c]%ip%[/id15a1d212c]; this will be substituted with the value of the same name inside the monitor (where you setup the script for that host).
Yes, thats it, but its not graphing anything, my graph options are -Data Type Counter, -Minium Value o, -Maximum Value 1.
I'v made some tests like give some urls at the parameters www.netmrg.pt
and the output of the gatherer is [code19a70b97e81][Dev: 0002] [Sub: 0049] [Mon: 0147] Sending '/usr/local/libexec/netmrg/check_http -H www.netmrg.pt'
[Dev: 0002] [Sub: 0049] [Mon: 0147] Value: Invalid host name 'www.netmrg.pt'
[/code19a70b97e81] and the Value / Data on netmrg is 0
and then with www.nokia.pt
[code19a70b97e81][Dev: 0002] [Sub: 0049] [Mon: 0147] Sending '/usr/local/libexec/netmrg/check_http -H www.nokia.pt'
[Dev: 0009] [Sub: 0053] [Mon: 0157] Value: HTTP ok: HTTP/1.1 200 OK - 1.021 second response time |time= 1.021
[/code19a70b97e81] and the Value / Data on netmrg is 1
what i wana do is if the response is "Value invalid host name" then the graph colum will be 0, if he response is "Value HTTP ok" the graph colum will be 1; Data / Values are working but he don't graph anything..
For your script test (check_http) you're collecting the error code and not the standard output, correct?
Noup, i'm collecting the standard output.