Freezes and issues after a week running
-
Yesterday we found some of our workstations could not access the Internet while others were fine. I found that I could not access the pfSense GUI. So we hooked up a monitor and keyboard to the router and found the console was repeating:
maxproc limit exceded by vid 0, please see… OVER and OVER
with
Approaching the limit on PV Entries, consider...
perproc or the the vm_pmap_pv_entry_max tunable... After so many linesWe had to power off the router manually. After restarting there were disk errors it corrected and all is well now.
It is an Atom 330, 2gb RAM, with a 1GB DOM and a dual Intel NIC PRO/1000 MT. Onboard NIC is not being used. It is version 1.2.3-RELEASE. I have 3 more routers with this same setup (except they use 2 Intel GT NICs, instead of 1 PRO/1000MT Dual Port) at different customer sites with no issues. This is our in house router. No VPNs are used and only two packages, Dashboard and NUT. I'm an IT guy but have little to no experience with BSD or really UNIX or LINUX.
Help! Thanks.
-
Ok, this happens every 24 hours now. I restart and it's great. Then it shows max pv entries need to be adjusted on the console. I have disabled DHCP and the DNS forwarder to 'lessen' the load but that made no difference.
Please help. Give me some advise.
-
Freezes of this type are almost always hardware related. Troubleshoot accordingly.
-
What packages do you have installed? Something is running away with processes.
-
@cmb:
What packages do you have installed? Something is running away with processes.
It worked well for 48 hours but did it again today. :(
The only packages I have installed are NUT and Dashboard. Should I try another Pro 1000MT? I have two more new in box in stock. Or is there a log/file perhaps that can show what is causing this?
Thanks
-
Do you have NAT reflection enabled?
-
@cmb:
Do you have NAT reflection enabled?
I checked and I did have NAT reflection enabled. I quickly disabled it as I meant for that to be disabled. I must have forgotten to re-disable it when testing a new port forward several weeks ago.
Could that have been the cause? Please explain if so.
-
Possibly, I have heard of a circumstance once where way too many nc instances get launched and makes that happen. No idea what circumstance that is, i haven't personally seen it.