Mysql have suddenly started regularly opening unsuccesful sockets -
i've desperately tried figure out what's happened here, haven't seen particular problem anywhere. i've 'inherited' (as in, not built of myself) management of database server (remote, in data warehouse, accessed ssh) php daemons running on linux server acting data crawlers, inserting , processing information in relatively steady stream mysql.
a couple of days ago, server crashed , came on again. logged in restarted mysql server , crawlers, thinking no more of it. day , half later, mysql server stopped working, , couldn't diagnose since couldn't log it, nor did respond "/etc/init.d/mysql stop" or varieties thereof. according log file, kept throwing errors regularly (once every 4 minutes , 16 seconds) , said had many file handlers open. when shut down crawlers, however, log in again, mysql kept throwing errors. checked lsof , showed lot of open sockets "can't identify protocol" error.
mysqld 28843 mysql 1990u sock 0,4 2856488 can't identify protocol
mysqld 28843 mysql 1989u sock 0,4 2857220 can't identify protocol
^thousands of these rows
i thought crawlers had done, , restarted mysql , failed sockets disappeared. surprised see mysql kept opening new ones, when crawlers weren't running. did regularly, 2 new failed sockets minute, regardless of whether crawlers active or not. increased maximum amount of filehandlers allowed mysql buy time, i'm looking diagnosis , permanent solution.
all descriptions of such errors (socket leaks) i've found on forums seems own software leaking, not closing sockets. seems mysql it, , there has been no change in of code when worked fine, server crash , restart.
any ideas?
Comments
Post a Comment