Our call center was completely down with thousands of sip_autodestruct messages in master node. During the incident, bascul commands didn't swap completely. Some resources were still hanging on master server. Even reboot commands didn't work on master server. So it was stuck in that failure state. I had to reboot from IPMI and then the service went to normal on secondary node. This happened last month as well. This was the second time.
When it happens, I saw there are over 450 calls in Vital dashboard. Actually in normal hour, it was less than 100 calls always.
Please help me check for this issue. I saw sip_autodestruct topic in many asterisk forum. But no one gave us the complete answer to avoid it.
It happened again. We saw many sip_autodestruct messages and channels are were releasing. We saw abnormal amount of concurrent calls in dashboard during the incident.