escape error
Hi,
I am running latest 3.1b, but I already noticed it back with 3.1a. But didnt yet had the time to look into.
After a connection loss to the master db in master/slave setup, slaves get DoS by this line:
So 19. Jun 22:11:23 CEST 2016 19.06.2016-22:11 - WARNING - Falsche Anfrage / Wrong QuerySQL-Query = INSERT INTO sys_log (server_id,datalog_id,loglevel,tstamp,message) VALUES (13, 0, 1, UNIX_TIMESTAMP(), 'DB::query -> reconnect -> 2003 (Can't connect to MySQL server on '10.0.0.1' (113))') -> 0 ()
Why I say DoS? The next seconds it repeats adding some backslashes:
So 19. Jun 22:11:23 CEST 2016 19.06.2016-22:11 - WARNING - Falsche Anfrage / Wrong QuerySQL-Query = INSERT INTO sys_log (server_id,datalog_id,loglevel,tstamp,message) VALUES (13, 0, 1, UNIX_TIMESTAMP(), 'Falsche Anfrage / Wrong QuerySQL = INSERT INTO sys_log (server_id,datalog_id,loglevel,tstamp,message) VALUES (13, 0, 1, UNIX_TIMESTAMP(), 'DB::query -> reconnect -> 2003 (Can\'t connect to MySQL server on \'10.0.0.1\' (113))') -> 0 ()') -> 0 ()
(Still the same second) Since the process runs every minute, after a few hours you got a few spare cron tasks running tail and taking your whole cpu time. I already saw a ~10gb cron.log.
Find the log of 10seconds attached, reproduce by taking master mysql offline for some time (reboot or upgrade for example) and restart it. bug