1
19.7 Legacy Series / Network outage randomly - Need help to investigate
« on: November 29, 2019, 02:54:34 pm »
Hi,
I used opnsense for few years now and I really like it !
I run a virtual machine on Proxmox (kvm) with 2vcpu and 2gb of ram, 10Gb hdd.
On this vm, I have 4 virtual interfaces with dedicated mac address and routing on the hoster network (ovh).
These interfaces are dedicated to haproxy to deliver web services, and 3 openvpn servers.
On the lan side, I have multiple vlan on the same interface. Each of this vlan is a /30 subnet where I configure a virtual server and an opnsense ip address for gateway.
It was working without any reboot for last 4 months. And, randomly last week, our services where not available anymore and we had to stop / restart the firewall.
Today, another outage and I tried to reboot directly the virtual machine without success, our services became available for 10 seconds. Then the firewall stopped to respond.
For troubleshoot, I checked at the arp table and found that every local ip had the same mac address.
I tried to stop the vm and to start it (cold boot) again, and miracle, everything seems to be fine and working again. I checked at the arp table and every local ip has a specific mac address now.
I think that the arp table was full, and everything dropped. The reboot did not flush the table, maybe because the table is directly reloaded in case of reboot ?
Please if anyone has any king of solution, investigation, or anything else ? I do not really know how to troubleshoot quickly this problem before it appears again ?
Thanks for your reply.
Regards,
I used opnsense for few years now and I really like it !
I run a virtual machine on Proxmox (kvm) with 2vcpu and 2gb of ram, 10Gb hdd.
On this vm, I have 4 virtual interfaces with dedicated mac address and routing on the hoster network (ovh).
These interfaces are dedicated to haproxy to deliver web services, and 3 openvpn servers.
On the lan side, I have multiple vlan on the same interface. Each of this vlan is a /30 subnet where I configure a virtual server and an opnsense ip address for gateway.
It was working without any reboot for last 4 months. And, randomly last week, our services where not available anymore and we had to stop / restart the firewall.
Today, another outage and I tried to reboot directly the virtual machine without success, our services became available for 10 seconds. Then the firewall stopped to respond.
For troubleshoot, I checked at the arp table and found that every local ip had the same mac address.
I tried to stop the vm and to start it (cold boot) again, and miracle, everything seems to be fine and working again. I checked at the arp table and every local ip has a specific mac address now.
I think that the arp table was full, and everything dropped. The reboot did not flush the table, maybe because the table is directly reloaded in case of reboot ?
Please if anyone has any king of solution, investigation, or anything else ? I do not really know how to troubleshoot quickly this problem before it appears again ?
Thanks for your reply.
Regards,