System crash, Reboot, Incorrect information in file: './mysql/host.frm'

[Eric Gumtow]

My virtual hosting company unexpectedly yanked the power on my server. I was using the mysql 5 command line at the time. After my system booted back up, I noticed mysql was down.

I tried to restart mysql. I got this:

/etc/init.d# ./mysql start
Starting MySQL database server: mysqld . . . . . . . . . . . . . . failed!

My syslog has this:

Sep 11 03:11:25 ejgumtow mysqld_safe[2381]: started
Sep 11 03:11:25 ejgumtow mysqld[2384]: 070911 3:11:25 InnoDB: Database was not shut down normally!
Sep 11 03:11:25 ejgumtow mysqld[2384]: InnoDB: Starting crash recovery.
Sep 11 03:11:25 ejgumtow mysqld[2384]: InnoDB: Reading tablespace information from the .ibd files...
Sep 11 03:11:25 ejgumtow mysqld[2384]: InnoDB: Restoring possible half-written data pages from the doublewrite
Sep 11 03:11:25 ejgumtow mysqld[2384]: InnoDB: buffer...
Sep 11 03:11:25 ejgumtow mysqld[2384]: 070911 3:11:25 InnoDB: Starting log scan based on checkpoint at
Sep 11 03:11:25 ejgumtow mysqld[2384]: InnoDB: log sequence number 0 44044.
Sep 11 03:11:25 ejgumtow mysqld[2384]: InnoDB: Doing recovery: scanned up to log sequence number 0 44044
Sep 11 03:11:25 ejgumtow mysqld[2384]: 070911 3:11:25 InnoDB: Started; log sequence number 0 44044
Sep 11 03:11:25 ejgumtow mysqld[2384]: 070911 3:11:25 [Note] Recovering after a crash using /var/log/mysql/mysql-bin
Sep 11 03:11:25 ejgumtow mysqld[2384]: 070911 3:11:25 [Note] Starting crash recovery...
Sep 11 03:11:25 ejgumtow mysqld[2384]: 070911 3:11:25 [Note] Crash recovery finished.
Sep 11 03:11:25 ejgumtow mysqld[2384]: 070911 3:11:25 [ERROR] /usr/sbin/mysqld: Incorrect information in file: './mysql/host.frm'
Sep 11 03:11:25 ejgumtow mysqld[2384]: 070911 3:11:25 [ERROR] /usr/sbin/mysqld: Incorrect information in file: './mysql/host.frm'
Sep 11 03:11:25 ejgumtow mysqld[2384]: 070911 3:11:25 [ERROR] Fatal error: Can't open and lock privilege tables: Incorrect information in file: './mysql/host.frm'
Sep 11 03:11:25 ejgumtow mysqld_safe[2395]: ended
Sep 11 03:11:40 ejgumtow /etc/init.d/mysql[2530]: 0 processes alive and '/usr/bin/mysqladmin --defaults-file=/etc/mysql/debian.cnf ping' resulted in
Sep 11 03:11:40 ejgumtow /etc/init.d/mysql[2530]: ^G/usr/bin/mysqladmin: connect to server at 'localhost' failed
Sep 11 03:11:40 ejgumtow /etc/init.d/mysql[2530]: error: 'Can't connect to local MySQL server through socket '/var/run/mysqld/mysqld.sock' (2)'
Sep 11 03:11:40 ejgumtow /etc/init.d/mysql[2530]: Check that mysqld is running and that the socket: '/var/run/mysqld/mysqld.sock' exists!

From which the important stuff seems to be:

Incorrect information in file: './mysql/host.frm'

I'm guessing that file is corrupt. I've googled and searched here. There are lots of disjoint solutions - none of which solve my problem.

Any suggestions? I haven't yet tried a backup, reinstall, restore. I'd like to avoid that if I can.

/# uname -a
Linux ejgumtow.com 2.6.11.12-xenU #5 Sat Jun 17 05:17:56 GMT 2006 i686 GNU/Linux

Thanks.