MySQL server PID file could not be found! after MySQL 5.5 update

Fred.

Verified User
Joined
Sep 5, 2009
Messages
220
Hello,

I've updated my MySQL 5.1 to 5.5 with custombuild an now i get
Code:
MySQL server PID file could not be found!

Any solution? my website is down at the moment.

I'm on CentOS 5.6
And the PID file is in /var/lib/mysql/

Thanks!
 
Last edited:
We had an issue with libaio missing on a server when upgrading it to 5.5. You could try and install it using yum.

# yum -y install libaio

And then try again. If it's still not solved, check your log files.
 
@Fred,

Where is the PID file kept in 5.1? Is it in the same location?

Jeff
 
I gave up for now...

MySQL deleted the pid file and can't upgrade it for MySQL 5.5

When i upload the pid file again and restart MySQL, the pid file is gone. :confused:
 
And what are you having in MySQL error log? What if you try with empty /etc/my.cnf?

p.s. I feel like trying MySQL 5.5. myself, as I have time )))
 
Last edited:
And what are you having in MySQL error log? What if you try with empty /etc/my.cnf?

p.s. I feel like trying MySQL 5.5. myself, as I have time )))


I can try with an empty /etc/my.cnf tomorrow at night, now it's too busy.
 
Last edited:
And what are you having in MySQL error log? What if you try with empty /etc/my.cnf?

p.s. I feel like trying MySQL 5.5. myself, as I have time )))

Removed the my.cnf and installed MySQL without any problems. :cool:
Now i need to set my custom config back in my.cnf and find out what was wrong. :p
 
If someone knows what's wrong.

This is my my.cnf,and it's not working on MySQL 5.5.11

Code:
[mysqld]
local-infile=0
safe-show-database
back_log = 50
#skip-innodb
max_connections = 500
key_buffer_size = 64M
myisam_sort_buffer_size = 64M
join_buffer_size = 1M
read_buffer_size = 1M
sort_buffer_size = 2M
table_cache = 4000
thread_cache_size = 256
wait_timeout = 200
connect_timeout = 10
tmp_table_size = 64M
max_heap_table_size = 64M
max_allowed_packet = 64M
net_buffer_length = 16384
max_connect_errors = 100000
thread_concurrency = 4
concurrent_insert = 2
table_lock_wait_timeout = 30
read_rnd_buffer_size = 786432
bulk_insert_buffer_size = 8M
query_cache_limit = 3M
query_cache_size = 64M
query_cache_type = 1
query_prealloc_size = 262144
query_alloc_block_size = 65536
range_alloc_block_size = 4096
transaction_alloc_block_size = 8192
transaction_prealloc_size = 4096
default-storage-engine = MyISAM
max_write_lock_count = 8
pid-file=/var/lib/mysql/server01.example.com.pid
socket          = /var/lib/mysql/mysql.sock

[mysqld_safe]
nice = -5
open_files_limit = 8192
log-error=/var/log/mysqld.log
pid-file=/var/lib/mysql/server01.example.com.pid

[mysqldump]
quick
max_allowed_packet = 16M

[myisamchk]
key_buffer = 64M
sort_buffer = 64M
read_buffer = 16M
write_buffer = 16M

I removed the pid file value in my.cnf but still not working.:confused:
 
Last edited:
Please tell us what you removed; in the event someone has a similar problem in the future, it will help.

Thanks.

Jeff
 
From my experience, some options that works in 5.1 but removed in 5.5 caused MySQL can't start. What I know for now is :

- safe-show-database (Removed in 5.5.3)
- skip-locking (Use "skip-external-locking" instead)

So, if you have above options, you will get PID problem. In case of Fred, I believe that because he has "safe-show-database" in my.cnf.

Code:
110429  4:39:15 [ERROR] /usr/sbin/mysqld: unknown option '--safe-show-database'
110429  4:39:15 [ERROR] Aborting

Also, be careful to use

- log_slow_queries=filename

It's deprecated since 5.1.29 and will remove in 5.6. You can use "slow_query_log" instead.
 
Last edited:
I came across this problem today in a new install on CentOS 6, immediately after installing the update to 5.5 from the default. Server not yet in use.

First I removed the skip-locking line from my my.cnf file, then I commented out the skip-innodb line from the same file, and saved my changes.

Then, in this order:

I stopped the mysql daemon,
I removed the two ib_logfile* files
I restarted mysql

and it's working now with 5.5

Does anyone know if that's good or bad (removing the skip-innodb line?

Jeff
 
You should use ignore_builtin_innodb instead with MySQL 5.5

This option causes the server to behave as if the built-in InnoDB is not present. It has these effects:

Other InnoDB options (including --innodb and --skip-innodb) will not be recognized and should not be used.

The server will not start if the default storage engine is set to InnoDB. Use --default-storage-engine to set the default to some other engine if necessary.

InnoDB will not appear in the output of SHOW ENGINES.

http://dev.mysql.com/doc/refman/5.5/en/innodb-parameters.html#option_mysqld_ignore-builtin-innodb
 
Back
Top