MySQL Bugs: #39820: Fail to start mysql server

Bug #39820	Fail to start mysql server
Submitted:	2 Oct 2008 17:38	Modified:	8 Nov 2008 9:11
Reporter:	Alina Alexeeva	Email Updates:
Status:	No Feedback	Impact on me:	None
Category:	MySQL Server: Installing	Severity:	S1 (Critical)
Version:	5.0.21	OS:	Linux
Assigned to:	Assigned Account	CPU Architecture:	Any
Tags:	cannot connect to the server

Description:
Have got repeated assertion failures even immidietly after the mysqld startup. Yesterday was my first day and i turned the computer on and started mysql server. tryed to log on but only received error: 'Can't connect to local MySQL server through socket '/var/run/mysqld/mysqld.sock' (2)'Check that mysqld is running and that the socket: '/var/run/mysqld/mysqld.sock' exists!. I added to file config.inc.php line
$cfg['Servers'][$i]['socket']='/var/run/mysqld/mysqld.sock'; (the same path as in my.cnf for variable 'socket')
and managed to connect through phpmyadmin, although attempts to connect through comand line gave me the sme error.
Today I turned on the computer and couldn't connect via cmd nor phpmyadmin. started investigating syslog. The same from today's and yesterday's log:
081001  9:03:35  InnoDB: Database was not shut down normally!
Oct  1 09:03:35 localhost mysqld[4626]: InnoDB: Starting crash recovery.
Oct  1 09:03:35 localhost mysqld[4626]: InnoDB: Reading tablespace information from the .ibd files...
Oct  1 09:03:35 localhost mysqld[4626]: InnoDB: Restoring possible half-written data pages from the doublewrite
Oct  1 09:03:35 localhost mysqld[4626]: InnoDB: buffer...
Oct  1 09:03:35 localhost mysqld[4626]: 081001  9:03:35  InnoDB: Starting log scan based on checkpoint at
Oct  1 09:03:35 localhost mysqld[4626]: InnoDB: log sequence number 24 82421643.
Oct  1 09:03:35 localhost mysqld[4626]: InnoDB: Doing recovery: scanned up to log sequence number 24 82421643
Oct  1 09:03:36 localhost mysqld[4626]: InnoDB: Page directory corruption: supremum not pointed to
Oct  1 09:03:36 localhost mysqld[4626]: 081001  9:03:36  InnoDB: Page dump in ascii and hex (16384 bytes):
Oct  1 09:03:36 localhost mysqld[4626]:  len 16384; hex 0000000000....
InnoDB: stored checksum 0, prior-to-4.0.14-form stored checksum 0
Oct  1 09:03:36 localhost mysqld[4626]: InnoDB: Page lsn 0 0, low 4 bytes of lsn at page end 0
Oct  1 09:03:36 localhost mysqld[4626]: InnoDB: Page number (if stored to page already) 0,
Oct  1 09:03:36 localhost mysqld[4626]: InnoDB: space id (if created with >= MySQL-4.1.1 and stored already) 0
Oct  1 09:03:36 localhost mysqld[4626]: InnoDB: Page directory corruption: supremum not pointed to
Oct  1 09:03:36 localhost mysqld[4626]: 081001  9:03:36  InnoDB: Page dump in ascii and hex (16384 bytes):
Oct  1 09:03:36 localhost mysqld[4626]:  len 16384; hex 000000...
09:03:36 localhost mysqld[4626]: Attempting backtrace. You can use the following information to find out
Oct  1 09:03:36 localhost mysqld[4626]: where mysqld died. If you see no messages after this, something went
Oct  1 09:03:36 localhost mysqld[4626]: terribly wrong...
Oct  1 09:03:36 localhost mysqld[4626]: Cannot determine thread, fp=0xbff344b8, backtrace may not be correct.
Oct  1 09:03:36 localhost mysqld[4626]: Stack range sanity check OK, backtrace follows:
Oct  1 09:03:36 localhost mysqld[4626]: 0x8189ef9
Oct  1 09:03:36 localhost mysqld[4626]: 0xffffe420
Oct  1 09:03:36 localhost mysqld[4626]: 0xbff35804
Oct  1 09:03:36 localhost mysqld[4626]: 0x830c42b
Oct  1 09:03:36 localhost mysqld[4626]: 0x82bb2e7
Oct  1 09:03:36 localhost mysqld[4626]: 0x82bc8b6
Oct  1 09:03:36 localhost mysqld[4626]: 0x82a8d48
Oct  1 09:03:36 localhost mysqld[4626]: 0x82a7452
Oct  1 09:03:36 localhost mysqld[4626]: 0x823443b
Oct  1 09:03:36 localhost mysqld[4626]: 0x822a52a
Oct  1 09:03:36 localhost mysqld[4626]: 0x81869b5
Oct  1 09:03:36 localhost mysqld[4626]: 0x818a457
Oct  1 09:03:36 localhost mysqld[4626]: 0xb7c79ea2
Oct  1 09:03:36 localhost mysqld[4626]: 0x8110bb1

How to repeat:
connecting to the service with InnoDB crashed previosly

Thank you for the report.

But version 5.0.27 is old and many bugs were fixed since. Current version is 5.0.67.

Also I believe your problem can be not result of a bug, because you said "connecting to the service with InnoDB crashed previosly".

Please check "Forcing InnoDB Recovery" page in the user manual at http://dev.mysql.com/doc/refman/5.0/en/forcing-recovery.html, try solution provided and if problem still exists, upgrade, try again and inform us if problem still exists.

Hello,
i've tried to reset forcing_innogb_recovery up to 6, but still have the same situation. Can you suggest anything exept upgrading. Or what can cause the failure

We're sorry, but the bug system is not the appropriate forum for asking help on using MySQL products.

Support on using our products is available both free in our forums at http://forums.mysql.com/ and for a reasonable fee direct from our skilled support engineers at http://www.mysql.com/support/

To identify if you are experiencing a bug in MySQL software we need you to use the last version of the MySQL. So I can not suggest here anything except upgrade right now.

I've installed MySQl 5.0.21 on Ubuntu 6.6 (for some reason for this OS there is now 5.0.67 available for downloading on the mysql site). Previoslly I uninstalled MySQL 5.0.41 (couldn't connect to the server mysqld got signal 11, repeated failures after mysqld starts) using
sudo aptitude --purge remove mysql-server-5.0 mysql-server
installed using
sudo apt-get install mysql-server
attempting to connect using
# /etc/init.d/mysql start
failed with message: /usr/bin/mysqladmin: connect to server at 'localhost' failed
error: 'Can't connect to local MySQL server through socket '/var/run/mysqld/mysqld.sock' (2)'
Check that mysqld is running and that the socket: '/var/run/mysqld/mysqld.sock' exists!

syslog:
mysqld_safe[5941]: started
localhost mysqld[5944]: 081007 12:26:10  InnoDB: Database was not shut down normally!
localhost mysqld[5944]: InnoDB: Starting crash recovery.
localhost mysqld[5944]: InnoDB: Reading tablespace information from the .ibd files...
localhost mysqld[5944]: InnoDB: Restoring possible half-written data pages from the doublewrite
localhost mysqld[5944]: InnoDB: buffer...
localhost mysqld[5944]: 081007 12:26:10  InnoDB: Starting log scan based on checkpoint at
localhost mysqld[5944]: InnoDB: log sequence number 0 36808.
localhost mysqld[5944]: InnoDB: Doing recovery: scanned up to log sequence number 0 43655
localhost mysqld[5944]: 081007 12:26:10  InnoDB: Starting an apply batch of log records to the database...
localhost mysqld[5944]: InnoDB: Progress in percents: 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 
Oct  7 12:26:11 localhost mysqld[5944]: InnoDB: Apply batch completed
Oct  7 12:26:11 localhost mysqld[5944]: 081007 12:26:11  InnoDB: Started; log sequence number 0 43655
localhost mysqld[5944]: 081007 12:26:11 [ERROR] /usr/sbin/mysqld: Can't find file: './mysql/host.frm' (errno: 13)
localhost mysqld[5944]: 081007 12:26:11 [ERROR] /usr/sbin/mysqld: Can't find file: './mysql/host.frm' (errno: 13)
localhost mysqld[5944]: 081007 12:26:11 [ERROR] Fatal error: Can't open and lock privilege tables: Can't find file: './mysql/host.frm' (errno: 13)
localhost mysqld_safe[5955]: ended
localhost /etc/init.d/mysql[6090]: 0 processes alive and '/usr/bin/mysqladmin --defaults-file=/etc/mysql/debian.cnf ping' resulted in
localhost /etc/init.d/mysql[6090]: ^G/usr/bin/mysqladmin: connect to server at 'localhost' failed
localhost /etc/init.d/mysql[6090]: error: 'Can't connect to local MySQL server through socket '/var/run/mysqld/mysqld.sock' (2)'
localhost /etc/init.d/mysql[6090]: Check that mysqld is running and that the socket: '/var/run/mysqld/mysqld.sock' exists!
localhost /etc/init.d/mysql[6090]:

Thank you for the feedback.

> I've installed MySQl 5.0.21 on Ubuntu 6.6 (for some reason for this OS there is now 5.0.67
available for downloading on the mysql site). 

You can use generic Linux packages. 5.0.21 is even older than originally used 5.0.27.

Regarding to the error posted: system tables are not installed.

No feedback was provided for this bug for over a month, so it is
being suspended automatically. If you are able to provide the
information that was originally requested, please do so and change
the status of the bug back to "Open".