pfsense reboot randomly on vmware
-
Cool. One of our devs is digging into this. He may reach out to you with a diagnostic kernel.
-
@stephenw10
maximum availability to support you and the community -
i think that for 2 Days 14 Hours 08 Minutes 08 Seconds the firt point of step 1. ( wireguard down for two days and see if problem backup job ) is done
today renable backup job and resubmit result . if crach disable vmxnet3 and enable e1000 -
system crashed after enable backup copy with vpn wireguard dump_1212_2023.txt
-
another crash Crash report begins. Anonymous mac.txt
-
Hmm, interesting.
Are you able to upgrade to 2.7.2 and confirm it also happens there?
-
I performed update before re-enabling backup jobs
-
@stephenw10
you thinh that is keepalive the problem ?now for all vpn is set to 25 second
-
It's shouldn't be. You should not be able to panic the kernel by changing the keep-alive time!
-
@stephenw10
so far it has not rebooted I am waiting for tonight's backups . I will update you tomorrow morning -
So that's using vmxnet3 NIC in pfSense 2.7.2 b ut with WireGuard disabled?
-
@stephenw10 said in pfsense reboot randomly on vmware:
So that's using vmxnet3 NIC in pfSense 2.7.2 b ut with WireGuard disabled?
i use vmxnet3 NIC in pfSense 2.7.2 with #3 vpn WireGuard . only one vpn have traffic copy backup
-
Ok great so WireGuard is still up but not passing traffic?
-
@stephenw10 said in pfsense reboot randomly on vmware:
Ok great so WireGuard is still up but not passing traffic?
wireguard has always been active. traffic always goes to the remote point, but when I activate the backup that has the remote point as a second copy pfsense crashes
-
Great we're trying to replicate it here.
-
if desired we can make a wireguard vpn pfsense to my router and give you an nfs endpoint to simulate the backup repository and test traffic
-
I was unable to make it panic. I set up a WireGuard tunnel between two 23.09.1 ESXi (7.0.3) VMs and tried to reproduce this in a couple ways. First with the iperf3 client/server on the firewalls themselves, then second with the iperf3 client/server on hosts behind the firewalls. Both were TCP tests, no NAT, which ran for over 8 hours each.
-
@security_sharezone If you're able to test this with iperf and still trigger it that then points to the ESXi version difference.
-
i have done various tests to isolate the problem and i think at this point it is the nfs outbound protocol. this is because from another true pfsense location always nfs but this time inbound i have no problems
-
@stephenw10
nothing continuos crash. you have idea ?
