2

So about a week ago, I finally figured out what was causing my server to continually crash.

After reviewing my mysqld.log I keep seeing this same error,

101210  5:04:32 [Warning] option 'max_join_size': unsigned value 18446744073709551615 adjusted to 4294967295

Here is a link to the bug report, http://bugs.mysql.com/bug.php?id=35346 someone recommend that you set the max_join_size vaule in my.cnf to 4M, and I did. I assumed this fixed the issue, and it was working for about a week with no issues until today...

I checked MySQL and the same error is now back,

101216 06:35:25  mysqld restarted
101216  6:38:15 [Warning] option 'max_join_size': unsigned value 18446744073709551615 adjusted to 4294967295
101216  6:38:15 [Warning] option 'max_join_size': unsigned value 18446744073709551615 adjusted to 4294967295
101216 06:40:42  mysqld ended

Anyone know how I can really fix this issue? I can't keep having mysql crash like this.

EDIT: I forgot to mention every time this happens I get an email from linode staying I have a high disk io rate

Your Linode, has exceeded the notification threshold (1000) for disk io rate by averaging 2483.68 for the last 2 hours.

2
  • How did you get MySQL? Installing it using yum? Compiled from source? What version are you using? (run mysqld --version)
    – Lekensteyn
    Dec 16, 2010 at 21:52
  • I installed via yum "mysql Ver 14.12 Distrib 5.0.77, for redhat-linux-gnu (i686) using readline 5.1"
    – mike
    Dec 16, 2010 at 22:18

3 Answers 3

1

For linode to email you about your IO threshold being exceeded for over two hours like that, I would think that something is causing heavy disk IO. Are you running low on memory or anything of the sort in the process? Often on VPS systems like Linode seeing high disk IO is due to low memory and swapping going on. OOMkiller may be being invoked which kills mysql. Does dmesg show any mention of oomkiller?

0

When you say the server is crashing - what exactly do you mean?

Do you mean that the system reboots or hangs(becomes unresponsive and must be reset/powercycled)?

Or do you mean only that MySQL is crashing - if so, do you mean that no MySQL processes are running or that all MySQL processes are unresponsive to queries?

Does anything appear on the system console when this occurs?

I am thinking that the MySQL warnings have nothing to do with the issue - my read of the bug report you list is that it is not something that causes MySQL to crash, yet alone crash the system.

0

The warning is just that - a warning. I don't think its your issue. Try to increase logging of mysql. Is it possible when you hit IO threshold limits Linode starts rejecting your IO requests, and this causes mysql to crash or end?

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .