Running an alter table to add an index was causing MySQL to crash with the following error:

ERROR 2013 (HY000): Lost connection to MySQL server during query

Was happening on 5.1.31. Pretty small table ~2m rows, so ended up just dumping the data and reimporting it and it fixed the problem. This was only happening on the production box, not in dev.

A

I was trying to add an index to a table with ~117M rows and was receiving the following error:

ERROR 1034 (HY000): Incorrect key file for table ‘frogs’

Moving my tmpdir to another drive with more space and rerunning the alter table solved the issue.

A

Ran into this one setting up a MySQL proxy, turns out i cant type 3406 and typed 3046. It produced the following when trying to connect:

ERROR 2003 (HY000): Can’t connect to MySQL server on ” (111)

Fixed it by using the correct port…

A

MySQL 5.1.26 and 5.1.28 (newer versions may also do this) seem to discontinue output during InnoDB recovery, you may see the following with no additional output for 30-40 minutes:

081105 20:06:29 mysqld_safe Starting mysqld daemon with databases from /mysecretdata/mysql

Followed by the usual warning:

InnoDB: Log scan progressed past the checkpoint lsn 57 793368317
081105 20:06:31  InnoDB: Database was not shut down normally!
InnoDB: Starting crash recovery.

Notice the timestamp is right after the first message but did not appear in the log until 30 minutes later. The output appears to be buffered somewhere and not written to the log. So if you are wondering why you arn’t seeing your InnoDB recovery progress, this could be the reason.

A