Dear Howtoforge, A master ISPconfig server went out of space for mysql databases resulting mysql service down (for a short period), the problem was when one of the slave servers went crazy and send 30000 msg to root witch is forwarded in /etc/aliases to my email address. The msg was: 26.08.2014-10:52 - WARNING - DB::query -> reconnect Can't connect to MySQL server on 'xx.xxx.net' (111) I only noticed 3 different dates in a warning msg: 10:52, 10:54, 10:56 , The second server didn't even complain. The system on a slave server is Debian wheezy. Any clue why would this happen? Best regards,
The mysql server must have been down so ispconfig was not able to connect to it and logged the errors for the reconnect attempts. I will take a look at the code so that we add something to prevent that so many duplicate error messages were sent.
is there any progress with this problem, for example in less than a minute of downtime it sends more than 2000 messages, so mailserver goes also down ... :/
Ok. WE added some changes to the git stable version which will be part of the next patch release. The db library was already limited to max 5 reconnect attempts. But under vers special circumstances it could happen that you geot more messages, (e.g. when the connection worked while the script read all websites into an array and then failed when tit tried to write the quota updates back to the db, then you would have got 10x the amount of websites that have as error messages as a reconnect attempt was done for each site.
