Experiencing instability on my Q2A

Question

It seems that my Q2A instance that used to run stable is now randomly going down every couple weeks or so with no explainable reason.

A system reboot seems to fix the problem. I'm running on Amazon EC2/Ubuntu - any thoughts on how to diagnose/check what might the issue might be?

[Update: It's happened quite a few times now. Every week or so, I check my Q2A instance and it seems to be having a problem with a database connection. I reset the server and it fixes itself.]

[Update again: I've found some stuff in the logs as follows:

root@ip-10-225-170-15:/var/log/mysql# cat error.log

131211 12:07:13 [Note] Plugin 'FEDERATED' is disabled.

131211 12:07:13 InnoDB: The InnoDB memory heap is disabled

131211 12:07:13 InnoDB: Mutexes and rw_locks use GCC atomic builtins

131211 12:07:13 InnoDB: Compressed tables use zlib 1.2.3.4

131211 12:07:13 InnoDB: Initializing buffer pool, size = 128.0M

InnoDB: mmap(137363456 bytes) failed; errno 12

131211 12:07:13 InnoDB: Completed initialization of buffer pool

131211 12:07:13 InnoDB: Fatal error: cannot allocate memory for the buffer pool

131211 12:07:13 [ERROR] Plugin 'InnoDB' init function returned error.

131211 12:07:13 [ERROR] Plugin 'InnoDB' registration as a STORAGE ENGINE failed.

131211 12:07:13 [ERROR] Unknown/unsupported storage engine: InnoDB

131211 12:07:13 [ERROR] Aborting

 

When I try to restart the `mysql' service:

 

root@ip-10-225-170-15:~# service mysql start

start: Job failed to start

 

When I restart the Amazon instance, it works fine. My googling reveals that it appears to be some kind of memory problem (probably due to the fact that I haven't allocated enough RAM to the server etc.)]

Comments

It's happened quite a few times now. Every week or so, I check my Q2A instance and it seems to be having a problem with a database connection. I reset the server and it fixes itself.

Answer 1

Try to allocate a swap space for your instance as described here :

http://www.prowebdev.us/2012/05/amazon-ec2-linux-micro-swap-space.html

and here :

http://stackoverflow.com/questions/10284532/amazon-ec2-mysql-aborting-start-because-innodb-mmap-x-bytes-failed-errno-12?rq=1

Comments

Thanks. Fairly sure that's the way to fix the issue. The odd thing is that it hasn't happened for the last 5 days (and I haven't applied the fix above yet.)

Weird.

Answer 2

It can't be a Q2A problem because Q2A runs under PHP which is stateless. So you might want to check your web server and/or database server logs for any reports of problems.

Comments

Thanks for the answer. I'll check out the logs and report back.