vicidial.org

Hi Guys

We had an unfortunate event where one of our diallers just died on us. Luckily we had other servers and decided to change all "Server IP"'s under "DID"s to point to other servers to ensure calls still route through to the correct servers while sorting our the other machine.
Everything else works, except for internal calls. When extension XXX tries to call ZZZ internally, a message saying "Not found" will appear.

Looking forward to hearing from the team.

Thanks

make sure that both extensions are on the same context

Hi there

Below is the asterisk CLI output from dialer 3.. (192.168.0.241 is the server that died)

This shows that the calls are still trying to reach the dead server even though it has been deactivated.

[Feb 24 08:24:34] -- Called dialer3:sc0MFVehaLmmXiu@192.168.0.241:4569/160
[Feb 24 08:24:37] -- Hungup 'IAX2/192.168.0.241:4569-4046'
[Feb 24 08:24:37] == Spawn extension (default, 192*168*000*241*160, 1) exited non-zero on 'SIP/cc336-00000026'
[Feb 24 08:24:37] -- Executing [h@default:1] DeadAGI("SIP/cc336-00000026", "agi://127.0.0.1:4577/call_log--HVcauses--PRI-----NODEBUG-----0-----CANCEL----------") in new stack
[Feb 24 08:24:37] -- AGI Script agi://127.0.0.1:4577/call_log--HVcauses ... ---------- completed, returning 0
[Feb 24 08:24:48] -- Executing [160@default---agent:1] Goto("SIP/cc336-00000027", "default|192*168*000*241*160|1") in new stack
[Feb 24 08:24:48] -- Goto (default,192*168*000*241*160,1)
[Feb 24 08:24:48] -- Executing [192*168*000*241*160@default:1] Dial("SIP/cc336-00000027", "IAX2/dialer3:sc0MFVehaLmmXiu@192.168.0.241:4569/160|55|oT") in new stack
[Feb 24 08:24:48] -- Called dialer3:sc0MFVehaLmmXiu@192.168.0.241:4569/160
[Feb 24 08:24:52] -- IAX2/192.168.0.241:4569-566 is circuit-busy
[Feb 24 08:24:52] NOTICE[2579]: chan_iax2.c:4082 __auto_congest: Auto-congesting call due to slow response
[Feb 24 08:24:52] -- Hungup 'IAX2/192.168.0.241:4569-566'
[Feb 24 08:24:52] == Everyone is busy/congested at this time (1:0/1/0)
[Feb 24 08:24:52] -- Executing [192*168*000*241*160@default:2] Hangup("SIP/cc336-00000027", "") in new stack
[Feb 24 08:24:52] == Spawn extension (default, 192*168*000*241*160, 2) exited non-zero on 'SIP/cc336-00000027'
[Feb 24 08:24:52] -- Executing [h@default:1] DeadAGI("SIP/cc336-00000027", "agi://127.0.0.1:4577/call_log--HVcauses--PRI-----NODEBUG-----16-----CONGESTION----------") in new stack
[Feb 24 08:24:52] -- AGI Script agi://127.0.0.1:4577/call_log--HVcauses ... ---------- completed, returning 0
dialer3*CLI>

Please help

Hoping to hear from the team the soonest

Nelius wrote:Hi there

Below is the asterisk CLI output from dialer 3.. (192.168.0.241 is the server that died)

This shows that the calls are still trying to reach the dead server even though it has been deactivated.

[Feb 24 08:24:34] -- Called dialer3:sc0MFVehaLmmXiu@192.168.0.241:4569/160
[Feb 24 08:24:37] -- Hungup 'IAX2/192.168.0.241:4569-4046'
[Feb 24 08:24:37] == Spawn extension (default, 192*168*000*241*160, 1) exited non-zero on 'SIP/cc336-00000026'
[Feb 24 08:24:37] -- Executing [h@default:1] DeadAGI("SIP/cc336-00000026", "agi://127.0.0.1:4577/call_log--HVcauses--PRI-----NODEBUG-----0-----CANCEL----------") in new stack
[Feb 24 08:24:37] -- AGI Script agi://127.0.0.1:4577/call_log--HVcauses ... ---------- completed, returning 0
[Feb 24 08:24:48] -- Executing [160@default---agent:1] Goto("SIP/cc336-00000027", "default|192*168*000*241*160|1") in new stack
[Feb 24 08:24:48] -- Goto (default,192*168*000*241*160,1)
[Feb 24 08:24:48] -- Executing [192*168*000*241*160@default:1] Dial("SIP/cc336-00000027", "IAX2/dialer3:sc0MFVehaLmmXiu@192.168.0.241:4569/160|55|oT") in new stack
[Feb 24 08:24:48] -- Called dialer3:sc0MFVehaLmmXiu@192.168.0.241:4569/160
[Feb 24 08:24:52] -- IAX2/192.168.0.241:4569-566 is circuit-busy
[Feb 24 08:24:52] NOTICE[2579]: chan_iax2.c:4082 __auto_congest: Auto-congesting call due to slow response
[Feb 24 08:24:52] -- Hungup 'IAX2/192.168.0.241:4569-566'
[Feb 24 08:24:52] == Everyone is busy/congested at this time (1:0/1/0)
[Feb 24 08:24:52] -- Executing [192*168*000*241*160@default:2] Hangup("SIP/cc336-00000027", "") in new stack
[Feb 24 08:24:52] == Spawn extension (default, 192*168*000*241*160, 2) exited non-zero on 'SIP/cc336-00000027'
[Feb 24 08:24:52] -- Executing [h@default:1] DeadAGI("SIP/cc336-00000027", "agi://127.0.0.1:4577/call_log--HVcauses--PRI-----NODEBUG-----16-----CONGESTION----------") in new stack
[Feb 24 08:24:52] -- AGI Script agi://127.0.0.1:4577/call_log--HVcauses ... ---------- completed, returning 0
dialer3*CLI>

Please help

Hoping to hear from the team the soonest

Here are some ideas -

Run rebuild of your config files. Admin -> Server -> Rebuild conf files

Also try to create two new extensions and see if they work.

Check your dial plan entries, you may have special routing settings, with hard coded IP there.

As Proper said "Check your dial plan entries, you may have special routing settings, with hard coded IP there"
If that is not the case then try to set this setting to "1" in Admin -> System Settings -> Generate Cross-Server Phone Extensions and submit. Let us know the update.

When a server dies and is replaced by another server (or the other servers are merely allowed to take over its functions), the only changes are:

1) Carriers that have inbound calls pointed to the dead server will fail inbound. Move all DIDs from the dead server to one of the other servers. Usually there is a failover system at the carrier to manage this automatically. If there is ... use it! That way a dead dialer will not cause inbound call disruption except for calls that were lilve at the time of death. Note that this is not done in Vicidial anywhere unless there are registration-based DIDs attached to the vicidial carrier entry. Usually DIDs are IP-based and configured at the Telco's website.

2) Phones (admin->phones) assigned to this server must be moved to another server. Doing so will not automatically change the values in the soft phone to point to the new server, so you'll need to change those manually at each phone.

After these have both been accomplished it's not a bad idea to modify all the servers to reload their configurations in all categories just to be sure nothing was overlooked.