It seems more like a linux problem, but I'm only seeing it on my SBCs using the 3CX distro, and it's killing my SBC fail-over.
The environment;
A pair of vmware SBCs configured to fail-over, everything is up to date.
A cloud PBX
Everything starts out looking good, both SBC are up and crm status shows everything is good, and my default route is shown on both SBCs. All the phones are running on sbc1.
I disconnect the interface (virtually) on sbc1 and everything fails-over very quickly to sbc2, the status of the SBCs flips in the management interface. All looks good.
Then I reconnect the interface of sbc1.
crm status on both shows that they can see each other, they are on the same LAN so that's good, but this error shows up on sbc1 and it seems that sbc2 is seeing sbc1 as on-line so it stops servicing the virtual IP of the SBC, they both show disconnected on the PBX.
Failed Actions:
* sbc-src-ip_stop_0 on [email protected] 'unknown error' (1): cal
l=9859, status=complete, exitreason='no matching route exists',
And sure enough, when I check the active routes on sbc1, the default route is gone. So the problem kind of makes sense, the two SBCs can see each other so sbc1 would be the active one, but it can't connect to the PBX because it has no default route.
I've check and recheck and it looks right on the interface, and I've tried it as both "gateway 192.168.0.1" and "up route add default gw 192.168.10.1" with identical results.
I understand why the route goes away when the interface is down, it's an invalid route, but it should come back. If I reboot the routing is back.
At this point I'm assuming I'm overlooking something simple, setting a default route on linux isn't all that complicated. I don't consider myself a general linux expert, I've used it for appliance type applications, and mostly CentOS, so maybe there is something I'm missing in Debian.
And help is greatly appreciated.
The environment;
A pair of vmware SBCs configured to fail-over, everything is up to date.
A cloud PBX
Everything starts out looking good, both SBC are up and crm status shows everything is good, and my default route is shown on both SBCs. All the phones are running on sbc1.
I disconnect the interface (virtually) on sbc1 and everything fails-over very quickly to sbc2, the status of the SBCs flips in the management interface. All looks good.
Then I reconnect the interface of sbc1.
crm status on both shows that they can see each other, they are on the same LAN so that's good, but this error shows up on sbc1 and it seems that sbc2 is seeing sbc1 as on-line so it stops servicing the virtual IP of the SBC, they both show disconnected on the PBX.
Failed Actions:
* sbc-src-ip_stop_0 on [email protected] 'unknown error' (1): cal
l=9859, status=complete, exitreason='no matching route exists',
And sure enough, when I check the active routes on sbc1, the default route is gone. So the problem kind of makes sense, the two SBCs can see each other so sbc1 would be the active one, but it can't connect to the PBX because it has no default route.
I've check and recheck and it looks right on the interface, and I've tried it as both "gateway 192.168.0.1" and "up route add default gw 192.168.10.1" with identical results.
I understand why the route goes away when the interface is down, it's an invalid route, but it should come back. If I reboot the routing is back.
At this point I'm assuming I'm overlooking something simple, setting a default route on linux isn't all that complicated. I don't consider myself a general linux expert, I've used it for appliance type applications, and mostly CentOS, so maybe there is something I'm missing in Debian.
And help is greatly appreciated.
