[OpenSIPS-Users] opensips segfault on mysql errors

[Chris Picton]

Hi all

I am running opensips 1.6.2 against a mysql cluster

On occasion, my sql nodes return temporary errors to opensips, which are
of the form 'please restart transaction'

An example of this would be:
ERROR:db_mysql:db_mysql_submit_query: driver error: Lock wait timeout
exceeded; try restarting transaction

This is generally followed by:
Dec 22 14:43:18 sbc-tpna1-01 /usr/sbin/opensips[22213]:
INFO:core:handle_sigs: child process 22272 exited by a signal 11
Dec 22 14:43:18 sbc-tpna1-01 /usr/sbin/opensips[22213]:
INFO:core:handle_sigs: core was generated
Dec 22 14:43:18 sbc-tpna1-01 /usr/sbin/opensips[22213]:
INFO:core:handle_sigs: terminating due to SIGCHLD
Dec 22 14:43:18 sbc-tpna1-01 /usr/sbin/opensips[22318]:
INFO:core:sig_usr: signal 15 received
Dec 22 14:43:18 sbc-tpna1-01 /usr/sbin/opensips[22310]:
INFO:core:sig_usr: signal 15 received
Dec 22 14:43:18 sbc-tpna1-01 /usr/sbin/opensips[22271]:
INFO:core:sig_usr: signal 15 received
Dec 22 14:43:18 sbc-tpna1-01 /usr/sbin/opensips[22251]:
INFO:core:sig_usr: signal 15 received
Dec 22 14:43:18 sbc-tpna1-01 /usr/sbin/opensips[22237]:
INFO:core:sig_usr: signal 15 received
Dec 22 14:43:18 sbc-tpna1-01 /usr/sbin/opensips[22224]:
INFO:core:sig_usr: signal 15 received
Dec 22 14:43:18 sbc-tpna1-01 /usr/sbin/opensips[22270]:
INFO:core:sig_usr: signal 15 received
Dec 22 14:43:18 sbc-tpna1-01 /usr/sbin/opensips[22248]:
INFO:core:sig_usr: signal 15 received
...
Dec 22 14:43:18 sbc-tpna1-01 /usr/sbin/opensips[22213]:
NOTICE:db_virtual:destroy: destroy module bla bla...


What should I be looking at to prevent this failure of sips when it sees
a database error?