Cacti has been up and running and working fine for some time. Today I tried to add interface polling to a new router and the intial query started hammering that router to a point where our DSL users speeds dropped to sub-dialup speeds. I couldn't get the interface query to stop so I rebooted the Cacti server itself.
This stopped the query, but now I can no longer poll any devices outside of 192.168.0.xxx, which is where the Cacti server is located. Under the monitoring screen, it says it can ping them but no SNMP data is returned. The few devices on the 192.168.0 network work fine, but we also have networks like 192.168.10, 192.168.11, and 12.180.180 that won't work.
I tried doing an SNMP walk on these sites and receive nothing back.
I downloaded a program called SNMP-Probe on my local PC, also in the 192.168.0 network, and was able to poll SNMP data from these other networks. Everything in the 192.168.0 has the same rights going through the firewall so if polling works for one device, it should work for them all. And nothing was changed on the firewall. So it seems to point to something wrong with the Cacti server but I'm not really sure where to start looking.
Can only poll devies on my local network...
Moderators: Developers, Moderators
- FlyingDuffman
- Posts: 25
- Joined: Tue Dec 26, 2006 12:18 pm
The first thing you should do if you assume that the firewalls don't block the SNMP traffic (UDP port 161) is to set the log to DEBUG Level, and take a moment to really read them...
An other thing you can do is to start an ethereal or something else, and check the SNMP traffic during the pool the of target devices. It seems to be a "brut" method, but believe me, it's usually a very fast method to find the problem
An other thing you can do is to start an ethereal or something else, and check the SNMP traffic during the pool the of target devices. It seems to be a "brut" method, but believe me, it's usually a very fast method to find the problem
- FlyingDuffman
- Posts: 25
- Joined: Tue Dec 26, 2006 12:18 pm
Thanks. I turned on debugging and saw lots of SNMP request timed out type errors. I had assumed from the monitoring tab that pings worked (since it said so on the monitoring tab) but found that when I tried to ping from the server itself, it did not work. I checked the network settings and found that all reference to my default gateway had disappeared. I entered the correct default gateway and all is well.
Who is online
Users browsing this forum: No registered users and 4 guests