linux - mysqld service stops once a day on ec2 server

Question

Environment Details:

Server: Amazon ec2 Linux
Web Server: Apache
Web Framework: Django with mod_wsgi

Following I have found in the mysql_err.log file.

The InnoDB memory heap is disabled
120823  3:21:40 InnoDB: Mutexes and rw_locks use GCC atomic builtins
120823  3:21:40 InnoDB: Compressed tables use zlib 1.2.3
120823  3:21:40 InnoDB: Using Linux native AIO
120823  3:21:41 InnoDB: Initializing buffer pool, size = 128.0M
InnoDB: mmap(137363456 bytes) failed; errno 12
120823  3:21:41 InnoDB: Completed initialization of buffer pool
120823  3:21:41 InnoDB: Fatal error: cannot allocate memory for the buffer pool
120823  3:21:41 [ERROR] Plugin 'InnoDB' init function returned error.
120823  3:21:41 [ERROR] Plugin 'InnoDB' registration as a STORAGE ENGINE failed.
120823  3:21:41 [ERROR] Unknown/unsupported storage engine: InnoDB
120823  3:21:41 [ERROR] Aborting

Looks like the system memory is not sufficient to allocate memory to buffer pool. Same error happens when I was using Amazon ec2 micro instance, So I moved to the small instance. It works fine for some days but now it is breaking once in a day again. Is there a permanent fix for that? I can move to medium instance but the issue is will that be fixed or not? Should I decrease the innodb_buffer_pool_size, what is the preferred size?

The result of cat /proc/meminfo is following (may be it will help):

MemTotal:        1697824 kB
MemFree:          125744 kB
Buffers:          109704 kB
Cached:           481408 kB
SwapCached:            0 kB
Active:          1212396 kB
Inactive:         266840 kB
Active(anon):     888192 kB
Inactive(anon):       76 kB
Active(file):     324204 kB
Inactive(file):   266764 kB
Unevictable:           0 kB
Mlocked:               0 kB
SwapTotal:             0 kB
SwapFree:              0 kB
Dirty:                 4 kB
Writeback:             0 kB
AnonPages:        888144 kB
Mapped:            15604 kB
Shmem:               144 kB
Slab:              63752 kB
SReclaimable:      53680 kB
SUnreclaim:        10072 kB
KernelStack:         800 kB
PageTables:        16436 kB
NFS_Unstable:          0 kB
Bounce:                0 kB
WritebackTmp:          0 kB
CommitLimit:      848912 kB
Committed_AS:    1417140 kB
VmallocTotal:   34359738367 kB
VmallocUsed:       10988 kB
VmallocChunk:   34359725168 kB
DirectMap4k:     1748992 kB
DirectMap2M:           0 kB

OS version (uname -a): Linux ip-10-246-134-149 3.2.21-1.32.6.amzn1.x86_64 #1 SMP Sat Jun 23 02:32:15 UTC 2012 x86_64 x86_64 x86_64 GNU/Linux

I checked the ps aux command whet the server has like only 15MB of memory left and these are the httpd process running at that time:

The result of free -m

total       used       free     shared    buffers     cached
Mem:          1657       1628         29          0          3         19
-/+ buffers/cache:       1605         51
Swap:          895        875         20

The result of ps aux

apache   21123  0.1  1.2 394652 20464 ?        S    19:35   0:06 /usr/sbin/httpd
apache   21146  0.1  1.2 394280 20796 ?        S    19:38   0:06 /usr/sbin/httpd
apache   21152  0.1  1.2 394284 21560 ?        S    19:38   0:05 /usr/sbin/httpd
apache   21155  0.2  1.4 396244 24528 ?        S    19:38   0:06 /usr/sbin/httpd
apache   21156  0.1  1.1 392552 20344 ?        S    19:38   0:06 /usr/sbin/httpd
apache   21157  0.1  1.1 394284 18884 ?        S    19:38   0:05 /usr/sbin/httpd
apache   21159  0.1  1.4 396200 25040 ?        S    19:38   0:06 /usr/sbin/httpd
apache   21161  0.1  1.2 394856 21724 ?        S    19:38   0:06 /usr/sbin/httpd
apache   21162  0.1  1.3 394864 22400 ?        S    19:38   0:06 /usr/sbin/httpd
apache   21163  0.1  1.3 394860 22204 ?        S    19:38   0:06 /usr/sbin/httpd
apache   21164  0.1  1.1 392560 19204 ?        S    19:38   0:06 /usr/sbin/httpd
apache   21165  0.1  1.3 394832 22280 ?        S    19:38   0:06 /usr/sbin/httpd
apache   21166  0.1  1.3 395276 22932 ?        S    19:38   0:06 /usr/sbin/httpd
apache   21172  0.2  1.4 396320 24820 ?        S    19:38   0:06 /usr/sbin/httpd
apache   21174  0.2  1.7 400672 29452 ?        S    19:39   0:06 /usr/sbin/httpd
apache   21178  0.1  1.4 400540 25304 ?        S    19:39   0:06 /usr/sbin/httpd
apache   21179  0.2  1.6 400580 27856 ?        S    19:39   0:06 /usr/sbin/httpd
apache   21184  0.1  1.7 400628 29320 ?        S    19:39   0:06 /usr/sbin/httpd
apache   21185  0.1  1.6 397944 27292 ?        S    19:39   0:05 /usr/sbin/httpd
apache   21186  0.1  1.5 397960 25648 ?        S    19:39   0:05 /usr/sbin/httpd
apache   21187  0.1  1.7 400576 29120 ?        S    19:39   0:06 /usr/sbin/httpd
apache   21191  0.1  1.4 400576 24400 ?        S    19:39   0:06 /usr/sbin/httpd
apache   21193  0.1  1.4 400536 24940 ?        S    19:39   0:05 /usr/sbin/httpd
apache   21194  0.1  1.5 400572 26096 ?        S    19:39   0:06 /usr/sbin/httpd
apache   21203  0.1  1.6 400580 28808 ?        S    19:39   0:05 /usr/sbin/httpd
apache   21206  0.1  1.7 400584 29732 ?        S    19:39   0:06 /usr/sbin/httpd
apache   21207  0.1  1.6 400576 27940 ?        S    19:39   0:06 /usr/sbin/httpd
apache   21224  0.1  1.2 400624 20768 ?        S    19:39   0:06 /usr/sbin/httpd
apache   21225  0.1  1.6 400576 28468 ?        S    19:39   0:05 /usr/sbin/httpd
apache   21226  0.1  1.6 400576 28048 ?        S    19:39   0:06 /usr/sbin/httpd
apache   21228  0.1  1.4 400572 23880 ?        S    19:39   0:06 /usr/sbin/httpd
apache   21237  0.1  1.5 400628 26124 ?        S    19:39   0:06 /usr/sbin/httpd
apache   21265  0.1  1.6 400536 28592 ?        S    19:39   0:06 /usr/sbin/httpd
apache   21276  0.1  1.2 400544 21456 ?        S    19:39   0:05 /usr/sbin/httpd
apache   21277  0.1  1.3 400624 22676 ?        S    19:39   0:05 /usr/sbin/httpd
apache   21278  0.1  1.6 400536 27360 ?        S    19:39   0:06 /usr/sbin/httpd
apache   21282  0.1  1.4 400612 24996 ?        S    19:39   0:06 /usr/sbin/httpd
apache   21292  0.1  1.4 400532 24780 ?        S    19:39   0:05 /usr/sbin/httpd
apache   21302  0.2  1.2 400540 21332 ?        S    19:39   0:06 /usr/sbin/httpd
apache   21303  0.1  1.3 400628 22228 ?        S    19:39   0:06 /usr/sbin/httpd
apache   21305  0.2  1.2 400536 21116 ?        S    19:39   0:06 /usr/sbin/httpd
apache   21306  0.1  1.3 400572 22380 ?        S    19:39   0:06 /usr/sbin/httpd
apache   21307  0.1  1.1 397956 20056 ?        S    19:39   0:05 /usr/sbin/httpd
apache   21308  0.1  1.2 400624 21520 ?        S    19:39   0:06 /usr/sbin/httpd
apache   21319  0.1  1.1 400540 19468 ?        S    19:39   0:05 /usr/sbin/httpd
apache   21320  0.1  1.3 400628 22712 ?        S    19:39   0:05 /usr/sbin/httpd
apache   21335  0.1  1.0 400540 17236 ?        S    19:39   0:05 /usr/sbin/httpd
apache   21336  0.1  1.3 400628 22188 ?        S    19:39   0:06 /usr/sbin/httpd
apache   21352  0.1  1.1 394276 18972 ?        S    19:40   0:04 /usr/sbin/httpd
apache   21356  0.1  1.1 394280 19028 ?        S    19:40   0:05 /usr/sbin/httpd
apache   21358  0.1  1.1 394280 19004 ?        S    19:40   0:05 /usr/sbin/httpd
apache   21361  0.2  0.7 400452 12632 ?        S    19:40   0:06 /usr/sbin/httpd
apache   21610  0.2  1.6 400536 27660 ?        S    19:46   0:06 /usr/sbin/httpd
apache   21643  0.2  1.3 400156 23272 ?        S    19:55   0:04 /usr/sbin/httpd
apache   21647  0.2  1.0 400544 17556 ?        S    19:57   0:05 /usr/sbin/httpd
apache   21654  0.2  1.5 400188 26884 ?        S    19:58   0:05 /usr/sbin/httpd
apache   21719  0.3  1.9 400192 32264 ?        S    20:14   0:03 /usr/sbin/httpd
apache   21725  0.2  2.0 400044 35340 ?        S    20:15   0:03 /usr/sbin/httpd
apache   21738  0.0  0.8 257648 13792 ?        S    20:26   0:00 /usr/sbin/httpd

Can any one one has an idea about it why there is so much httpd process??

Answer 1

Use 50% of available RAM to test:

You can decrease the innodb_buffer_pool_size very low to see if it helps:

#/etc/my.cnf 
innodb_buffer_pool_size = 1M

A rule of thumb is to set innodb_buffer_pool_size to 50% of available RAM for your low memory testing. This means you start the server and everything except MySQL InnoDB. See how much RAM you have. Then use 50% of that for InnoDB.

To try many low-memory settings at once:

• http://paragasu.wordpress.com/2008/12/02/very-low-memory-mysql-5-mycnf-configuration/

A more likely culprit is whatever else is on that server, such as a webserver.

Apache?

Are you using Apache and/or another webserver? If so, try to decrease its RAM usage. For example in Apache conf, consider low RAM settings like these:

StartServers 1
MinSpareServers 1
MaxSpareServers 5
MaxClients 5

And cap the requests like this:

MaxRequestsPerChild 300

Then restart Apache.

mod_wsgi:

If you're using Apache with mod_python, switch to Apache with mod_wsgi.

Pympler:

If it's still happening, possibly your Django is steadily growing. Try Django memory profiling with Pympler:

• http://www.rkblog.rk.edu.pl/w/p/profiling-django-object-size-and-memory-usage-pympler/

SAR:

Your report of once-per-day failures, then once-per-week failures, could point to some kind of cron job running daily or weekly. For example, perhaps there's a batch process that takes up a lot of RAM, or a database dump, etc.

To track RAM use and look for RAM spikes in the hour before MySQL dies, take a look at SAR, which is a great tool: http://www.thegeekstuff.com/2011/03/sar-examples/