Hello everyone!



This is my first time posting on this mailing list, a Mysql developer referred me here in hopes that I can find some solutions to our fairly unique situation. I’m pretty new to FBSD and Mysql so please bare with me.



I am the part owner of a pretty large website – www.midnitechallenge.com – our current member base is around 30,000 and about 400-500 concurrent players are online during peak hours, and anywhere from 100-200 during off-peak. We are running the current system:







Dual P3 2.1gig



2gig RAM



2 x 18gig SCSI drives



FBSD-stable 4.2.3



Mysql 3.23.54a



Innodb database – for row level locking (no transactions are actually being used though)







The site relies heavily on PHP/mysql and is currently hitting an average of about 400 queries/sec during peak times. Load averages are around 1.5 with 60-75% CPU usage devoted to mysqld.







The problem we are experiencing is that every 6-8hours, mysqld will suddenly ramp from 60-75% CPU usage to 99.02% within around 20-30secs and then hang. The only way to fix it is the kill -9 the process. Then everything will be fine and dandy for another 6-8hours.. its pretty much like clock work. I am by no means familiar enough with FSB or Mysql to tweak the settings or configs nor diagnose the cause of the problem, although I have played with certain settings in my.cnf and httpd.conf with no success.







I’ve noticed in the .err file, there are a lot of the strange entries right before mysqld restarts on its own, or hangs. You can view a recent err file here:







www.midnitechallenge.com/temp/mysql-error.err







I’ve actually been at the console and witnessed mysqld hang a few times and it would happen just as I described. I’ve found no mention of this kind of symptom on any websites and am at a complete loss as to how to resolve this. A few people have suggested that it may be a query that is hanging it, but I do not know how to confirm this. Another has suggested that this is an Innodb problem and I feel that this may be true, since prior to switching the tables over to Innodb, we didn’t experience this symptom. What gets me, is that everywhere I’ve read and everyone I’ve talked to, the system we are running should have no problems at all handling the load that we currently have, yet the numbers that we see and cpu usage don’t seem to support this – so the conclusion is that something isn’t right.







I’ve read a few things on http://jeremy.zawodny.com and have tried changing some of the configs but haven’t had any success. One thing I’m thinking of trying is installing Linux Threads to see if this resolves the issues, but have not proceeded to do so since there seems to be new problems that may be introduced as mentioned on Jeremy’s site..







I’m not sure what other information I can offer that may be of assistance – but any suggestions on how we can resolve this would be greatly appreciated.







Crescent Kao



Director of Marketing and Sales



www.corruptinteractive.com







PS- I just subscribed to the list and didn’t get a confirmation or anything so, not sure if I will receive and replies.










_________________________________________________________________
Protect your PC - get McAfee.com VirusScan Online http://clinic.mcafee.com/clinic/ibuy/campaign.asp?cid=3963


---------------------------------------------------------------------
Before posting, please check:
http://www.mysql.com/manual.php (the manual)
http://lists.mysql.com/ (the list archive)

To request this thread, e-mail <[EMAIL PROTECTED]>
To unsubscribe, e-mail <[EMAIL PROTECTED]>
Trouble unsubscribing? Try: http://lists.mysql.com/php/unsubscribe.php

Reply via email to