Take the 2-minute tour ×
WordPress Development Stack Exchange is a question and answer site for WordPress developers and administrators. It's 100% free, no registration required.

I have a Wordpress+Buddypress installation which is regularly going down - the number of HTTPD on the server and the server CPU load both peak suddenly and the server fails to recover. A restart of the web service is required.

I'm looking for troubleshooting advice.

Edit: Moved from the comment in here to make the goal clear.

I'm not getting any useful information out of the error log, even when enabling debug mode in Wordpress. I have tidied up a lot of sloppy code in plugins and themes which were throwing PHP notices, to no effect.

I have defined ERRORLOGFILE to log MU errors, but nothing has been logged there as yet. I have tried using Debug Bar for Wordpress with SQL logging enabled, but this has not shown anything useful - as far as I can see it does not log db transactions to file, and the Debug Bar only shows db queries for one site at a time. We have over 3000 sites running in this installation.

Other information: we are using WP Super Cache, using mod_rewrite. I have noticed quite a few errors in the error log to do with Wordpress trying to add or remove cache files which don't exist. I don't know if this is normal, or relevant.

Wordpress and Buddypress are the most recent versions as of writing: 3.5.1 / 1.6.4

I am looking for more tips in pinpointing the cause of the issue. I have started gradually disabling recently activated plugins, but this is not going to be practical for all plugins on the live site.

Any ideas? Thanks.

share|improve this question
    
Welcome to WPSE. About your question) Without any actual information it's less than guessing... sorry, but this won't work out. –  kaiser Mar 7 '13 at 18:28
    
Hi. I'm not expecting anyone to hand me a solution at this point. I'm looking for troubleshooting advice. –  Lemmy Mar 7 '13 at 20:08
    
Maybe you want to extend it with some part of your error log? Could be that we can interpret something that wasn't clear for you. –  kaiser Mar 7 '13 at 21:59
    
@kaiser: Regarding your edit of the post - I think the question was already quite clear from the text in the post "I am looking for more tips in pinpointing the cause of the issue". Anyway, thanks for taking the time to answer and adding in the TL:DR version. –  Lemmy Mar 8 '13 at 13:04

2 Answers 2

Check the Apache processes. I had this same sort of issue a while back and essentially what was happening was that the requests were being made faster than the server could handle it and at that point every slowed to a crawl. This was leaving hundreds of Apache processes running and waiting for the next thing that never came.The server load was up to 15-20 on the CPU. I restarted Apache and the problem was solved... Until it got overloaded again....

I hope that helps somehow. :)

share|improve this answer
1  
Oh, and many of those connections were from one ip. So I blocked it and it seemed to improve, but not 100%... Changing to nginx now. –  user28589 Mar 8 '13 at 10:11
    
Thanks - the culprit for me was the caching plugin - causing similar spikes, I think through mod_rewrite rules not working correctly, and possibly due to a plugin setting to rebuild cache after each new post. –  Lemmy Mar 8 '13 at 13:11

The problem was due to the use of the WP Super Cache plugin. Switching away from the mod_rewrite mode to use PHP mode, and deselecting some options in the plugin which the docs suggest might cause CPU spikes, has resulted in stable server performance.

share|improve this answer

Your Answer

 
discard

By posting your answer, you agree to the privacy policy and terms of service.

Not the answer you're looking for? Browse other questions tagged or ask your own question.