We use mysql to update status of a call using the built in mysql functions. We currently have been trying to switch to a different SIP provider for outgoing calls but once we hit a higher call volumes things start to go bad. Mysql deadlocks records and asterisk spits out a bunch of errors. Can’t figure out what side is the problem exactly except to think its the provider since it doesnt seem to happen on the original provider.
MySQL:
From what we can tell, the main source of sql trouble is around this deadlocking " Query SELECT LAST_INSERT_ID() AS last FROM table" which is autoincrement feature. It causes these errors:
WARNING[29079] app_mysql.c: mysql_real_connect(mysql,server,user,pass,DB,…) failed(2003): Can’t connect to MySQL server on ‘server’ (111)
WARNING[28766] app_mysql.c: aMYSQL_query: mysql_query failed. Error: MySQL server has gone away
RTP issue:
Getting tones of these errors once its starts going bad but mostly the last 4 errors.
ERROR[22211] res_rtp_asterisk.c: RTCP SR transmission error to 4.55.38.70:6257, rtcp halted Operation not permitted
ERROR[22211] res_rtp_asterisk.c: RTCP SR transmission error to 208.93.47.133:6699, rtcp halted Operation not permitted
ERROR[27394] acl.c: Cannot create socket
ERROR[27394] chan_sip.c: Got SDP but have no RTP session allocated.
WARNING[27394] res_rtp_asterisk.c: Unable to allocate RTP socket: Too many open files
WARNING[27394] chan_sip.c: Autodestruct on dialog ‘39e967dd6619b35941c7ce1477da4794@SERVERIP:5060’ with owner in place (Method: BYE)
Are the calls failing at a high rate causing mysql to execute all the failed queries at once locking up mysql? There are a couple queries that happen on the h extension in the context.
Any help on what exactly the RTP issues are having problems with?