MySQL Load after the last few updates

[rhenderson]

I have had MySQL load topping out slowing the server as of the Current 16150
2007-08-10 12:09:45 update.

It takes a few hours for the load to build. I have applied all current updates thinking it would help it out. I had some custom settings in my.cnf and removed them in case there was some conflict with the later versions.

I had turned on CPHulk and thought maybe it was causing the problems but still no luck. Somewhere between 4 and 6 hours it starts. I have not been able to track down a specific process causing it but generally I have to restart it to make things move fast enough to check out everything.

Any suggestions or if someone has the same thing happening share what you have done ot eliminate the problem.

Thanks!!

 

[nyjimbo]

Have you checked the "Show mysql processess" in WHM when this starts to happen?. If anything is going to clue you in that would be it. Also see if you have statistics configured to run during that time. You might be experiencing a load from several things and mysql is getting bogged down. If this is during a prime-time or heavy business time you might want to disable stats from running then to help drop the load

 

[rhenderson]

Strace

I did a strace on the pid and got:

clone(child_stack=0xb19e34c4, flags=CLONE_VM|CLONE_FS|CLONE_FILES|CLONE_SIGHAND|CLONE_THREAD|CLONE_SYSVSEM|CLONE_SETTLS|CLONE_PARENT_SETTID|CLONE_CHILD_CLEARTID|CLONE_DETACHED, parent_tidptr=0xb19e3be8, {entry_number:6, base_addr:0xb19e3ba0, limit:1048575, seg_32bit:1, contents:0, read_exec_only:0, limit_in_pages:1, seg_not_present:0, useable:1}, child_tidptr=0xb19e3be8) = 29205
select(13, [10 12], NULL, NULL, NULL)   = 1 (in [12])
fcntl64(12, F_SETFL, O_RDWR|O_NONBLOCK) = 0
accept(12, {sa_family=AF_FILE, path="×"}, [2]) = 56
fcntl64(12, F_SETFL, O_RDWR)            = 0
getsockname(56, {sa_family=AF_FILE, path="/var/lib/mysql"}, [28]) = 0
fcntl64(56, F_SETFL, O_RDONLY)          = 0
fcntl64(56, F_GETFL)                    = 0x2 (flags O_RDWR)
fcntl64(56, F_SETFL, O_RDWR|O_NONBLOCK) = 0
setsockopt(56, SOL_IP, IP_TOS, [8], 4)  = -1 EOPNOTSUPP (Operation not supported)
time(NULL)                              = 1187206671

Obviously something is a miss from the (Operation not supported)

 

[nyjimbo]

Obviously something is a miss from the (Operation not supported)

But wouldnt that more likely cause a dump or segfault and not just an increase in load.

 

[rhenderson]

It is up right now and the processes show:

 Main >> SQL Services >> Show MySQL Processes

MySQL Process List
Id User Host db Command Time State Info 

114 eximstats localhost eximstats Sleep 24  NULL 

1436 eximstats localhost eximstats Sleep 830  NULL 

1721 afou_agbook1 localhost afou_agbook1 Query 341 Copying to tmp table select x.*, y.p_filename, y.width, y.height, z.comments from book_data x left join book_pics y on (x 

1746 eximstats localhost eximstats Sleep 255  NULL 

1802 afou_agbook1 localhost afou_agbook1 Query 123 Copying to tmp table select x.*, y.p_filename, y.width, y.height, z.comments from book_data x left join book_pics y on (x 

1803 afou_agbook1 localhost afou_agbook1 Query 117 Copying to tmp table select x.*, y.p_filename, y.width, y.height, z.comments from book_data x left join book_pics y on (x 

1847 eximstats localhost eximstats Sleep 2  NULL 

1848 root localhost NULL Query 0 NULL SHOW PROCESSLIST

 

[rhenderson]

I removed that guestbook above, we will see what happens. Write permission were wrong and it was an old PHP 4 script so time will tell.

 

[rhenderson]

It took care of it. It is just strange that script was installed back in March and just now started give the server problems. Thanks for your help.