Cacti + pfsense vm = angry
-
we setup our install of cacti to poll all of our virtual machines interfaces every 5minutes, this includes our pfsense routers,
primary router on 2.0 rc3 with load kernel panics a few times a day when polled by cacti via snmp.
while at the same time the backup router with the same snmp polling has no problem(but again it has no load)
snmp claims there are something like 132 interfaces it responds with although we are only graphing the lan and wan
im wondering if there is a conflict with the internal graphing pfsense does? the crashes occured at very suspicious times (prob same time as a cron task), the load on the servers and the network is low when the panic occurs. but it occurs at
12 midnight
12 noon
3pmon the dot, makes me think a conflict may be the reason, i was suspecting it might be some type of dos attack, but i turned off the cacti graphs and the router has run fine for 24hours.
yikes!
attached are the screen shots of the kernel panics.
any suggestion on how to graph the pfsense interfaces with cacti without kernel panics would be great!
-
There must be something specific that's triggering it. I monitor multiple pfSense boxes with cacti and I have zero problems.
first, update to 2.0-RELEASE, and then if it's still problematic, provide quite a bit more info about the setup you have there (type of VM, hypervisor, resources allocated, type of NICs, etc)
-
There must be something specific that's triggering it. I monitor multiple pfSense boxes with cacti and I have zero problems.
Any specific templates which you are using for CACTI to grab pfsense status except interface traffic one?
-
No, just standard unix host template. I'm not graphing anything exotic.
If you want to graph pf bits, check out the pf mib
http://files.pfsense.org/jimp/BEGEMOT-PF-MIB.txt
