Description of problem: Timeout when starting the mysql daemon Version-Release number of selected component (if applicable): How reproducible: Steps to Reproduce: 1. 2. 3. Actual results: Expected results: Additional info:
Type of machine? Does it happen on more than one machine? What is the precise error it gives you, and after how much time? Anything showing up in logs?
Two steps to a resolution: Check if /usr/bin/mysqladmin -uUNKNOWN_MYSQL_USER ping runs from the command line while mysql is running. If it does not, add an anonymous user with no permissions to your server, change the UNKNOWN_MYSQL_USER to portion of the command to match your new user, and try running it again. eg /usr/bin/mysqladmin -u anonymous ping Your results would look something like this: /usr/bin/mysqladmin -uUNKNOWN_MYSQL_USER ping /usr/bin/mysqladmin: connect to server at 'localhost' failed error: 'Access denied for user 'UNKNOWN_MYSQL_US'@'localhost' /usr/bin/mysqladmin -u anonymous ping mysqld is alive If you've gotten to this point, update /etc/init.d/mysqld and change the line 41 to your new command. OK, now scroll down, and change line 79 in the restart section from sleep 2 to sleep 5 These two changes should solve things for you. Please post a note if they do. The proper error text for others searching is: Timeout error occurred trying to start MySQL Daemon. Hopefully they try searching in the comments. Your report is a touch light on details. The more details you include the more likely others will find your bug if they have a similar one or be able to help you if they solved it.
I am seeing the same timeout error with mysql-server-4.1.7-4.RHEL4.1.i386. Changing the timeout will not help, as the server doesn't start up: The following is from /var/log/mysqld.log 050320 15:45:29 mysqld started 050320 15:45:29 [ERROR] bdb: unable to initialize mutex: Function not implemented 050320 15:45:29 [ERROR] bdb: process-private: unable to initialize environment lock: Function not implemented 050320 15:45:30 InnoDB: Started; log sequence number 0 43634 050320 15:45:30 [ERROR] Can't init databases 050320 15:45:30 [ERROR] Aborting 050320 15:45:30 InnoDB: Starting shutdown... 050320 15:45:33 InnoDB: Shutdown completed; log sequence number 0 43634 050320 15:45:33 [Note] /usr/libexec/mysqld: Shutdown complete 050320 15:45:33 mysqld ended
error10 Yours is a different issue. In this one mysql starts but the ping fails, and the timeout is too short. What is your enviroment? Are you running on redhat binaries? I'd recommend posting a new bug (if not a dup) and including more information on your enviroment if you are, including any issues around forcing updates etc. You might also consider skip-bdb in /etc/my.conf [mysqld] as a short term fix. Please do specify more of your enviroment however, you will find folks who are more familiar with all the various factors that can create a bug reluctant to help unless they can narrow the range of possabilities somewhat.
I doubt that "timeout too short" is the issue. Either the mysql daemon is failing to start at all (as Michael suggests) or the daemon is fine but the method the init script uses to check whether the daemon is alive is wrong. There is a known problem with the latter bit --- see bug #142328 for instance --- but I believe this is fixed in mysql-4.1.7-5.RHEL4.1 and up. The initial report in this thread is so content-free that I can't tell whether that's the issue or not. I'm interested in Michael's comment though. That seems to indicate a pthread failure, which would not be mysql's fault in particular, but it would suggest a glibc problem. I've not seen it in my own testing though. Michael, what glibc package do you have exactly?
Closing this bug for lack of sufficient information to investigate the problem (or problems?) If you still see it with current RHEL4 packages, feel free to reopen.