logical Linking-Error in status.cgi
Posted: Fri Aug 27, 2010 6:47 am
We are using Nagios Core 3.2.1
Maybe the way we use nagios is a bit unusual, but we found a logical error in the Link-structure of the status.cgi, that should be general.
Since solving the problem seems to be done by reprogramming and recompiling the code, i can´t do it myself. Perhaps this can be fixed in the next release
Background:
Our Hostcheck is done by ping to the hosts IP.
We are also monitoring VPN Routers.
We enter its public ip as the address, so the hostcheck is a ping to the public ip to check if the router is online.
Then we have _tunnel-ip with the private ip, where we make a pingtest to check if the tunnel is working.
If the variable address is wrong or icmp on the external interface is blocked, the host is down and the service "ping to the private IP" is ok.
In this case in the navigation bar unter "Problems" we see no error at "Service Status Details For All Hosts", but at "Host Status Totals" we have one host shown as down.
When clicking on "Down" the list is empty, because the filter in the link excludes Hosts that have Servicestatus = OK.
I think, the reason is, that the programmers thought, that the service of a host can not be OK if the host is down. But that is obviously wrong.
/status.cgi?host=all&servicestatustypes=28&hoststatustypes=4
So the link should be
/status.cgi?host=all&servicestatustypes=31&hoststatustypes=4
or
/status.cgi?host=all&hoststatustypes=4
Or am I mistaking? Is there a fast way for me to fix this exept changing and recompiling the code?
best regards
Alexander Broemmelsiek
Maybe the way we use nagios is a bit unusual, but we found a logical error in the Link-structure of the status.cgi, that should be general.
Since solving the problem seems to be done by reprogramming and recompiling the code, i can´t do it myself. Perhaps this can be fixed in the next release
Background:
Our Hostcheck is done by ping to the hosts IP.
We are also monitoring VPN Routers.
We enter its public ip as the address, so the hostcheck is a ping to the public ip to check if the router is online.
Then we have _tunnel-ip with the private ip, where we make a pingtest to check if the tunnel is working.
Code: Select all
define host {
use generic-host-24x7
host_name hostname
address 12.34.56.78
_tunnel-ip 192.168.123.45
hostgroups pingtest_tunnel-24x7
}In this case in the navigation bar unter "Problems" we see no error at "Service Status Details For All Hosts", but at "Host Status Totals" we have one host shown as down.
When clicking on "Down" the list is empty, because the filter in the link excludes Hosts that have Servicestatus = OK.
I think, the reason is, that the programmers thought, that the service of a host can not be OK if the host is down. But that is obviously wrong.
/status.cgi?host=all&servicestatustypes=28&hoststatustypes=4
So the link should be
/status.cgi?host=all&servicestatustypes=31&hoststatustypes=4
or
/status.cgi?host=all&hoststatustypes=4
Or am I mistaking? Is there a fast way for me to fix this exept changing and recompiling the code?
best regards
Alexander Broemmelsiek