Honestly I only skimmed through the utils site. But, I will look into setting up a web interface when I have time to find the destructions on what to do. It's pointless to use SSH in my opinion...more of a pain to watch.
Till, I am using W3 Total Cache. Mark, any recommendations on a page outlining installing/setting up collectl output to a website? Not seeing it...
Here's the latest from tuning-primer.sh for Mysql: Code: SLOW QUERIES The slow query log is enabled. Current long_query_time = 10 sec. You have 14077 out of 16432161 that take longer than 10 sec. to complete Your long_query_time seems to be fine BINARY UPDATE LOG The binary update log is NOT enabled. You will not be able to do point in time recovery See http://dev.mysql.com/doc/refman/5.0/en/point-in-time-recovery.html WORKER THREADS Current thread_cache_size = 8 Current threads_cached = 7 Current threads_per_sec = 0 Historic threads_per_sec = 0 Your thread_cache_size is fine MAX CONNECTIONS Current max_connections = 200 Current threads_connected = 4 Historic max_used_connections = 72 The number of used connections is 36% of the configured maximum. Your max_connections variable seems to be fine. INNODB STATUS Current InnoDB index space = 64 K Current InnoDB data space = 64 K Current InnoDB buffer pool free = 95 % Current innodb_buffer_pool_size = 8 M Depending on how much space your innodb indexes take up it may be safe to increase this value to up to 2 / 3 of total system memory MEMORY USAGE Max Memory Ever Allocated : 381 M Configured Max Per-thread Buffers : 899 M Configured Max Global Buffers : 58 M Configured Max Memory Limit : 957 M Physical Memory : 3.87 G Max memory limit seem to be within acceptable norms KEY BUFFER Current MyISAM index space = 131 M Current key_buffer_size = 16 M Key cache miss rate is 1 : 1381 Key buffer free ratio = 0 % You could increase key_buffer_size It is safe to raise this up to 1/4 of total system memory; assuming this is a dedicated database server. QUERY CACHE Query cache is enabled Current query_cache_size = 32 M Current query_cache_used = 28 M Current query_cache_limit = 2 M Current Query cache Memory fill ratio = 88.43 % Current query_cache_min_res_unit = 4 K However, 101653 queries have been removed from the query cache due to lack of memory Perhaps you should raise query_cache_size MySQL won't cache query results that are larger than query_cache_limit in size SORT OPERATIONS Current sort_buffer_size = 2 M Current read_rnd_buffer_size = 256 K Sort buffer seems to be fine JOINS Current join_buffer_size = 2.00 M You have had 6 queries where a join could not use an index properly You should enable "log-queries-not-using-indexes" Then look for non indexed joins in the slow query log. If you are unable to optimize your queries you may want to increase your join_buffer_size to accommodate larger joins in one pass. Note! This script will still suggest raising the join_buffer_size when ANY joins not using indexes are found. OPEN FILES LIMIT Current open_files_limit = 1210 files The open_files_limit should typically be set to at least 2x-3x that of table_cache if you have heavy MyISAM usage. Your open_files_limit value seems to be fine TABLE CACHE Current table_cache value = 500 tables You have a total of 391 tables You have 452 open tables. The table_cache value seems to be fine TEMP TABLES Current max_heap_table_size = 80 M Current tmp_table_size = 80 M Of 49835 temp tables, 39% were created on disk Perhaps you should increase your tmp_table_size and/or max_heap_table_size to reduce the number of disk-based temporary tables Note! BLOB and TEXT columns are not allow in memory tables. If you are using these columns raising these values might not impact your ratio of on disk temp tables. TABLE SCANS Current read_buffer_size = 128 K Current table scan ratio = 1207 : 1 read_buffer_size seems to be fine TABLE LOCKING Current Lock Wait ratio = 1 : 1884 You may benefit from selective use of InnoDB.
Mark, my LOAD AVG was at 3.56 when I looked at TOP randomly. Started Collectl. Here's the output below up until the load avg was down to 1.10. Code: # <--------CPU--------><-----------Memory-----------><----------Disks-----------><----------Network----------> #Time cpu sys inter ctxsw Free Buff Cach Inac Slab Map KBRead Reads KBWrit Writes KBIn PktIn KBOut PktOut 00:53:40 33 16 61 1138 162M 669M 1G 1G 129M 2G 0 0 0 0 0 6 0 4 00:53:41 36 9 95 587 162M 669M 1G 1G 129M 2G 60 4 0 0 3 38 20 39 00:53:42 44 21 52 1487 164M 669M 1G 1G 129M 2G 4 1 3760 29 4 49 4 46 00:53:43 0 0 14 72 164M 669M 1G 1G 129M 2G 0 0 0 0 1 7 2 6 00:53:44 25 8 22 94 164M 669M 1G 1G 129M 2G 0 0 0 0 1 10 5 8 00:53:45 81 36 67 911 164M 669M 1G 1G 129M 2G 0 0 3228 18 2 16 1 9 00:53:46 39 22 34 1319 164M 669M 1G 1G 129M 2G 0 0 52 8 4 30 4 25 00:53:47 0 0 30 74 164M 669M 1G 1G 129M 2G 0 0 0 0 2 18 7 23 00:53:48 2 1 5 61 306M 669M 1G 1G 129M 2G 0 0 0 0 1 9 1 7 00:53:49 0 0 23 76 306M 669M 1G 1G 129M 2G 0 0 0 0 0 5 0 1 00:53:50 0 0 20 88 306M 669M 1G 1G 129M 2G 0 0 0 0 2 20 2 18 00:53:51 19 1 55 169 306M 669M 1G 1G 128M 2G 12 3 216 9 1 7 0 3 00:53:52 50 30 39 61 306M 669M 1G 1G 128M 2G 8 2 0 0 3 44 35 42 00:53:53 5 1 77 178 306M 669M 1G 1G 128M 2G 0 0 0 0 1 19 12 16 00:53:54 1 0 27 72 306M 669M 1G 1G 128M 2G 0 0 0 0 3 46 3 35 00:53:55 50 3 24 175 306M 669M 1G 1G 128M 2G 0 0 0 0 1 15 5 9 00:53:56 3 0 147 233 306M 669M 1G 1G 128M 2G 56 4 0 0 2 30 27 31 00:53:57 1 0 86 104 307M 669M 1G 1G 128M 2G 4 1 3452 22 6 75 82 93 00:53:58 11 0 35 104 306M 669M 1G 1G 128M 2G 0 0 104 12 4 40 9 32 00:53:59 50 28 88 200 306M 669M 1G 1G 128M 2G 0 0 0 0 2 14 1 6 00:54:00 33 8 36 196 306M 669M 1G 1G 128M 2G 4 1 0 0 4 47 5 45 00:54:01 50 15 10 962 306M 669M 1G 1G 128M 2G 0 0 0 0 2 16 6 13 # <--------CPU--------><-----------Memory-----------><----------Disks-----------><----------Network----------> #Time cpu sys inter ctxsw Free Buff Cach Inac Slab Map KBRead Reads KBWrit Writes KBIn PktIn KBOut PktOut 00:54:02 36 18 98 1210 306M 669M 1G 1G 128M 2G 0 0 0 0 1 12 1 7 00:54:03 0 0 42 99 306M 669M 1G 1G 128M 2G 0 0 3672 17 5 63 41 61 00:54:04 0 0 31 72 306M 669M 1G 1G 128M 2G 0 0 0 0 1 15 6 17 00:54:05 42 4 66 244 289M 669M 1G 1G 128M 2G 0 0 0 0 2 21 1 9 00:54:06 64 21 42 140 305M 669M 1G 1G 128M 2G 0 0 0 0 4 44 3 37 00:54:07 17 9 38 76 306M 669M 1G 1G 128M 2G 0 0 24 5 2 20 9 16 00:54:08 3 0 45 101 306M 669M 1G 1G 128M 2G 0 0 3352 21 2 21 2 15 00:54:09 50 19 22 1335 306M 669M 1G 1G 128M 2G 0 0 112 6 2 12 2 11 00:54:10 17 10 7 590 306M 669M 1G 1G 128M 2G 0 0 0 0 1 8 2 6 00:54:11 13 1 23 90 306M 669M 1G 1G 128M 2G 0 0 0 0 0 3 1 3 00:54:12 1 0 251 92 306M 669M 1G 1G 128M 2G 4 1 3668 230 2 17 2 17 00:54:13 1 0 151 169 306M 669M 1G 1G 128M 2G 0 0 492 61 3 39 47 50 00:54:14 0 0 32 94 306M 669M 1G 1G 128M 2G 0 0 220 7 2 29 2 22 00:54:15 1 0 44 88 306M 669M 1G 1G 128M 2G 0 0 0 0 1 14 1 12 00:54:16 0 0 78 190 306M 669M 1G 1G 128M 2G 0 0 0 0 3 34 9 37 00:54:17 0 0 37 80 306M 669M 1G 1G 128M 2G 0 0 0 0 3 37 3 36 00:54:18 0 0 38 97 307M 669M 1G 1G 128M 2G 0 0 0 0 2 28 2 28 00:54:19 1 0 162 235 306M 669M 1G 1G 128M 2G 4 1 144 10 3 30 11 25 00:54:20 39 15 99 112 306M 669M 1G 1G 128M 2G 0 0 88 8 7 83 146 142 00:54:21 67 27 125 153 306M 669M 1G 1G 128M 2G 176 2 0 0 5 75 316 225 00:54:22 41 23 120 87 306M 669M 1G 1G 128M 2G 0 0 32 2 5 51 168 129 00:54:23 0 0 105 109 306M 669M 1G 1G 128M 2G 0 0 0 0 7 90 254 187 # <--------CPU--------><-----------Memory-----------><----------Disks-----------><----------Network----------> #Time cpu sys inter ctxsw Free Buff Cach Inac Slab Map KBRead Reads KBWrit Writes KBIn PktIn KBOut PktOut 00:54:24 0 0 107 128 306M 669M 1G 1G 128M 2G 0 0 0 0 4 64 221 159 00:54:25 1 0 148 91 306M 669M 1G 1G 128M 2G 4 1 3560 28 4 54 152 120 00:54:26 0 0 88 65 306M 669M 1G 1G 128M 2G 0 0 0 0 6 83 317 222 00:54:27 7 0 125 135 306M 669M 1G 1G 128M 2G 4 1 4 1 3 40 167 120 00:54:28 50 22 24 119 305M 669M 1G 1G 128M 2G 0 0 0 0 6 77 188 137 00:54:29 19 7 160 128 305M 669M 1G 1G 128M 2G 0 0 0 0 2 21 39 40 00:54:30 1 0 105 106 305M 669M 1G 1G 128M 2G 52 1 0 0 9 123 357 257 00:54:31 0 0 54 88 305M 669M 1G 1G 128M 2G 0 0 3456 25 4 66 191 134 00:54:32 0 0 31 63 305M 669M 1G 1G 128M 2G 0 0 0 0 1 14 1 7 00:54:33 0 0 38 80 306M 669M 1G 1G 128M 2G 0 0 0 0 2 22 2 13 00:54:34 32 7 30 163 305M 669M 1G 1G 128M 2G 0 0 108 7 2 18 7 17 00:54:35 47 27 40 526 305M 669M 1G 1G 128M 2G 0 0 3184 22 2 12 1 8 00:54:36 0 0 32 120 305M 669M 1G 1G 128M 2G 0 0 0 0 1 15 2 10 00:54:37 0 0 71 90 305M 669M 1G 1G 128M 2G 416 5 0 0 1 19 6 18 00:54:38 0 0 36 61 305M 669M 1G 1G 128M 2G 0 0 0 0 4 47 257 182 00:54:39 0 0 21 81 305M 669M 1G 1G 128M 2G 0 0 0 0 1 17 176 123 00:54:40 0 0 11 45 305M 669M 1G 1G 128M 2G 0 0 0 0 1 13 5 11 00:54:41 39 5 37 163 305M 669M 1G 1G 128M 2G 0 0 112 8 0 7 0 3 00:54:42 100 62 36 65 305M 669M 1G 1G 128M 2G 4 1 16 3 3 34 30 32 Curiously, my swap space is on hda which is the main hdd. I have an hdd which is a tarball backup (I create tarballs monthly of /var/www/ and so on). I'm thinking I should reconfigure it so that hdb has the swap. That should fix up some I/O right? It's SATA so I mean it would allow the system to use I/O on the primary hdd for primary purposes. It would then let the secondary hdd handle the swap. Right?
Any opinion on my question above??? Reposted: Curiously, my swap space is on hda which is the main hdd. I have an hdd which is a tarball backup (I create tarballs monthly of /var/www/ and so on). I'm thinking I should reconfigure it so that hdb has the swap. That should fix up some I/O right? It's SATA so I mean it would allow the system to use I/O on the primary hdd for primary purposes. It would then let the secondary hdd handle the swap. Right? Reason I'm reposting is that the datacenter techs believe that to be negligible. But, if it ends up swapping I figured it wouldn't be negligible at that point?
Do you use any php-accelerator? especially when only couple sites get most of hits using those will have very good results. I have been on your seat, and i used APC and it was very sufficient. In debian install is very easy and configuration is childs play. APC can give 90% more power to Apache&php in frequently accessed sites. Feel free to contact via PM if you want.
I installed xcache and memcache for PHP a couple of days ago. APC is another type then? Anyone feel free to comment on the above swap question!
I think this (Alternative PHP Cache) is the APC referred to above. I will look into adding this to my server as well.
XCache has improved performance for sure. I'm going to let them run for another day or so, and then switch to APC and test that out.
I was doing some more reading on this and it seems that of the 3 accelerators, APC is about right in the middle performance wise but comes out way on top stability and future support wise. http://stackoverflow.com/questions/930877/apc-vs-eaccelerator-vs-xcache http://2bits.com/articles/benchmarking-apc-vs-eaccelerator-using-drupal.html http://2bits.com/articles/benchmark...hes-apc-eaccelerator-and-xcache-compared.html http://www.scribd.com/doc/88689/apcfacebook Apparently facebook uses APC and also contributes code back to the project which pretty much settles it for me. Install was only 2 steps (OK 3 if you count restarting Apache) Code: apt-get install php-apc Code: vi /etc/php5/conf.d/apc.ini add the following to apc.ini Code: extension=apc.so apc.enabled=1 apc.shm_size=30 Restart apache and done.
These are reasons why we are using it too. Also PHP6 will have accelerator build-in, and based on APC.
crypted: i wanted to throw in my 2 cents when it came to your mysql usage. If you're having high traffic on a few wordpress sites, I'm guessing you're having large load in mysql too. Can you try increasing your key_buffer_size from 16M to something closer to your index size...like 128M (now that you're at 4G Ram) and report back after a restart and 2 days usage? I like to start higher and drop it down if it's overkill. My other thought is your temp tables. Poor queries (and any query running over fulltext searches), will end up in temp tables which are tables written to disk. A major drain on resources. Decreasing your slow query time from 10 sec to 5 sec will probably show more of those queries so you can analyze and tweak them.
Yeah, I've been tweaking it every day and half. What's 44 hours vs. 48 hours really? So anyway, I've changed many-a-setting with SQL and it's gotten a lot better. I tweaked it again this afternoon; I'll report back with those results the next day or two. But, RAM + PHP CACHING + SQL TWEAKS = 0.36 avg load.
Falko: good tutorial to have. I've been using PRIMER and TUNER. I realize they are different in their suggestions, but having both does help get a good sensibility for what's needed. I really recommend using caching on high PHP websites after this. SQL was the next issue and it's pretty well resolved I think.
Today, we had a few minute lockup but I wasn't at home to observe it. My CPUWAIT hit 198. http://monitor.derekgordon.com/munin/derekgordon.com/my.derekgordon.com-cpu.html My processes average 198, however at the time of this occurrence the processes was over 360. Swap was utilized during this time. First time since the RAM increase to 4GB that this happened. http://monitor.derekgordon.com/munin/derekgordon.com/my.derekgordon.com-swap.html MYSQL has been tweaked and should be pretty good. Here's the output of the tuner.sh: Code: MySQL Version 5.0.51a-24+lenny4-log x86_64 Uptime = 3 days 0 hrs 57 min 33 sec Avg. qps = 384 Total Questions = 100897230 Threads Connected = 10 Server has been running for over 48hrs. It should be safe to follow these recommendations To find out more information on how each of these runtime variables effects performance visit: http://dev.mysql.com/doc/refman/5.0/en/server-system-variables.html Visit http://www.mysql.com/products/enterprise/advisors.html for info about MySQL's Enterprise Monitoring and Advisory Service SLOW QUERIES The slow query log is enabled. Current long_query_time = 10 sec. You have 55472 out of 100898023 that take longer than 10 sec. to complete Your long_query_time seems to be fine BINARY UPDATE LOG The binary update log is NOT enabled. You will not be able to do point in time recovery See http://dev.mysql.com/doc/refman/5.0/en/point-in-time-recovery.html WORKER THREADS Current thread_cache_size = 8 Current threads_cached = 3 Current threads_per_sec = 0 Historic threads_per_sec = 0 Your thread_cache_size is fine MAX CONNECTIONS Current max_connections = 200 Current threads_connected = 9 Historic max_used_connections = 76 The number of used connections is 38% of the configured maximum. Your max_connections variable seems to be fine. INNODB STATUS Current InnoDB index space = 64 K Current InnoDB data space = 64 K Current InnoDB buffer pool free = 98 % Current innodb_buffer_pool_size = 64 M Depending on how much space your innodb indexes take up it may be safe to increase this value to up to 2 / 3 of total system memory MEMORY USAGE Max Memory Ever Allocated : 649 M Configured Max Per-thread Buffers : 1.07 G Configured Max Global Buffers : 232 M Configured Max Memory Limit : 1.30 G Physical Memory : 3.87 G Max memory limit seem to be within acceptable norms KEY BUFFER Current MyISAM index space = 134 M Current key_buffer_size = 16 M Key cache miss rate is 1 : 395 Key buffer free ratio = 15 % You could increase key_buffer_size It is safe to raise this up to 1/4 of total system memory; assuming this is a dedicated database server. QUERY CACHE Query cache is enabled Current query_cache_size = 150 M Current query_cache_used = 54 M Current query_cache_limit = 24 M Current Query cache Memory fill ratio = 36.20 % Current query_cache_min_res_unit = 4 K MySQL won't cache query results that are larger than query_cache_limit in size SORT OPERATIONS Current sort_buffer_size = 2 M Current read_rnd_buffer_size = 256 K Sort buffer seems to be fine JOINS Current join_buffer_size = 3.00 M You have had 4 queries where a join could not use an index properly You should enable "log-queries-not-using-indexes" Then look for non indexed joins in the slow query log. If you are unable to optimize your queries you may want to increase your join_buffer_size to accommodate larger joins in one pass. Note! This script will still suggest raising the join_buffer_size when ANY joins not using indexes are found. OPEN FILES LIMIT Current open_files_limit = 1810 files The open_files_limit should typically be set to at least 2x-3x that of table_cache if you have heavy MyISAM usage. Your open_files_limit value seems to be fine TABLE CACHE Current table_cache value = 800 tables You have a total of 386 tables You have 498 open tables. The table_cache value seems to be fine TEMP TABLES Current max_heap_table_size = 375 M Current tmp_table_size = 375 M Of 153649 temp tables, 32% were created on disk Perhaps you should increase your tmp_table_size and/or max_heap_table_size to reduce the number of disk-based temporary tables Note! BLOB and TEXT columns are not allow in memory tables. If you are using these columns raising these values might not impact your ratio of on disk temp tables. TABLE SCANS Current read_buffer_size = 128 K Current table scan ratio = 1749 : 1 read_buffer_size seems to be fine TABLE LOCKING Current Lock Wait ratio = 1 : 1666 You may benefit from selective use of InnoDB. I would venture a guess this is to do with rampant apache/fcgi processes. However, I changed FCGI to 1 child as Till suggested. Have a raw.gz collectl for that day. Honestly, I can't figure out how to get it to display information in the terminal worth anything. It just floods everything and screws up my terminal connection. Help appreciated.
Apparently it's all related to fcgid. The processes continue to spawn even though I have the Max set to 1 as Till suggested. My load has been a constant 2-3 avg the past two hours. Namely it's php-cgi (fastcgi) that's doing the eating. Code: mysql 15381 6.7 2.9 416660 120856 pts/1 Sl 20:40 3:24 /usr/sbin/mysqld --basedir=/usr --datadir=/var/lib/mysql --user=mys root 15382 0.0 0.0 5040 668 pts/1 S 20:40 0:00 logger -p daemon.err -t mysqld_safe -i -t mysqld postfix 16961 0.0 0.0 38900 2288 ? S 20:50 0:00 pickup -l -t fifo -u -c root 17097 0.0 0.4 330864 17196 ? Ss 20:53 0:00 /usr/sbin/apache2 -k start root 17098 0.0 0.1 39188 6916 ? S 20:53 0:00 vlogger (access log) www-data 17099 0.0 0.1 170056 5320 ? S 20:53 0:00 /usr/sbin/apache2 -k start www-data 17100 0.0 0.3 332128 12396 ? S 20:53 0:02 /usr/sbin/apache2 -k start web5 17105 5.5 1.8 272388 73556 ? R 20:53 2:06 /usr/bin/php-cgi -d open_basedir=/var/www/clients/client1/web5/web: web5 17137 6.8 1.7 270096 71096 ? S 20:53 2:34 /usr/bin/php-cgi -d open_basedir=/var/www/clients/client1/web5/web: web11 17184 0.1 1.3 266256 55208 ? S 20:54 0:02 /usr/bin/php-cgi -d open_basedir=/var/www/clients/client1/web11/web web1 17212 0.2 2.4 284140 101260 ? S 20:54 0:05 /usr/bin/php-cgi -d open_basedir=/var/www/clients/client1/web1/web: web7 17562 0.0 1.3 263636 55780 ? S 20:55 0:01 /usr/bin/php-cgi -d open_basedir=/var/www/clients/client1/web7/web: web8 17688 6.3 1.9 279888 81276 ? S 20:57 2:08 /usr/bin/php-cgi -d open_basedir=/var/www/clients/client1/web8/web: www-data 18153 0.0 0.3 332516 12564 ? S 21:00 0:00 /usr/sbin/apache2 -k start web11 18279 0.0 1.3 263892 54904 ? S 21:03 0:01 /usr/bin/php-cgi -d open_basedir=/var/www/clients/client1/web11/web web8 18869 12.4 1.8 271824 74228 ? S 21:06 3:08 /usr/bin/php-cgi -d open_basedir=/var/www/clients/client1/web8/web: web8 18875 17.7 1.8 271808 73208 ? S 21:06 4:27 /usr/bin/php-cgi -d open_basedir=/var/www/clients/client1/web8/web: web8 18878 11.1 1.8 272296 73916 ? R 21:06 2:47 /usr/bin/php-cgi -d open_basedir=/var/www/clients/client1/web8/web: root 19546 0.0 0.0 10560 2196 pts/1 T 21:13 0:00 nano /home/dingle/collectl.output.log web9 20052 0.0 1.1 261224 47412 ? S 21:17 0:00 /usr/bin/php-cgi -d open_basedir=/var/www/clients/client1/web9/web: web9 20053 0.0 1.1 260956 47264 ? S 21:17 0:00 /usr/bin/php-cgi -d open_basedir=/var/www/clients/client1/web9/web: amavis 20106 0.0 0.2 192480 8524 ? S 21:18 0:00 amavisd (virgin child) web29 20124 0.0 1.3 267280 56724 ? S 21:18 0:00 /usr/bin/php-cgi -d open_basedir=/var/www/clients/client1/web29/web web29 20125 0.0 1.4 267272 57292 ? S 21:18 0:00 /usr/bin/php-cgi -d open_basedir=/var/www/clients/client1/web29/web www-data 20174 0.0 0.3 332104 12348 ? S 21:19 0:00 /usr/sbin/apache2 -k start www-data 20179 0.0 0.3 332136 12372 ? S 21:19 0:00 /usr/sbin/apache2 -k start www-data 20184 0.0 0.3 332104 12324 ? S 21:19 0:00 /usr/sbin/apache2 -k start www-data 20598 0.0 0.3 332504 12496 ? S 21:21 0:00 /usr/sbin/apache2 -k start www-data 20599 0.0 0.3 332128 12300 ? S 21:21 0:00 /usr/sbin/apache2 -k start web5 20609 1.7 1.9 276432 77288 ? S 21:21 0:10 /usr/bin/php-cgi -d open_basedir=/var/www/clients/client1/web5/web: web5 20610 3.8 1.9 276444 77320 ? S 21:21 0:22 /usr/bin/php-cgi -d open_basedir=/var/www/clients/client1/web5/web: nobody 20954 0.0 0.0 28184 2712 ? S Oct25 1:22 /usr/local/sbin/mydns -b web8 21141 14.2 1.7 271808 73140 ? S 21:26 0:45 /usr/bin/php-cgi -d open_basedir=/var/www/clients/client1/web8/web: web8 21146 6.5 1.7 271816 73088 ? S 21:26 0:20 /usr/bin/php-cgi -d open_basedir=/var/www/clients/client1/web8/web: web8 21149 23.8 1.7 271812 73088 ? S 21:26 1:14 /usr/bin/php-cgi -d open_basedir=/var/www/clients/client1/web8/web: web8 21150 1.1 1.7 271812 73032 ? S 21:26 0:03 /usr/bin/php-cgi -d open_basedir=/var/www/clients/client1/web8/web: web8 21151 2.2 1.7 271808 73080 ? S 21:26 0:06 /usr/bin/php-cgi -d open_basedir=/var/www/clients/client1/web8/web: web8 21155 12.1 1.7 271808 73140 ? S 21:26 0:37 /usr/bin/php-cgi -d open_basedir=/var/www/clients/client1/web8/web: web8 21159 16.8 1.7 271808 73128 ? S 21:26 0:51 /usr/bin/php-cgi -d open_basedir=/var/www/clients/client1/web8/web: www-data 21207 0.0 0.2 332064 11988 ? S 21:27 0:00 /usr/sbin/apache2 -k start www-data 21215 0.0 0.3 332112 12260 ? S 21:27 0:00 /usr/sbin/apache2 -k start www-data 21216 0.0 0.2 332112 12168 ? S 21:27 0:00 /usr/sbin/apache2 -k start www-data 21221 0.0 0.3 332388 12500 ? S 21:27 0:00 /usr/sbin/apache2 -k start postfix 21223 0.0 0.1 106484 6112 ? S 21:27 0:00 smtpd -n smtp -t inet -u -c -o stress postfix 21226 0.0 0.0 38900 2244 ? S 21:27 0:00 anvil -l -t unix -u -c web8 21238 4.1 1.9 277936 81056 ? S 21:27 0:09 /usr/bin/php-cgi -d open_basedir=/var/www/clients/client1/web8/web: web8 21244 2.6 1.9 277820 80944 ? S 21:27 0:05 /usr/bin/php-cgi -d open_basedir=/var/www/clients/client1/web8/web: amavis 21362 0.0 0.2 192480 8380 ? S 21:29 0:00 amavisd (virgin child) postfix 21546 0.0 0.0 38900 2224 ? S 21:30 0:00 showq -t unix -u -c www-data 21797 0.0 0.2 331920 11788 ? S 21:30 0:00 /usr/sbin/apache2 -k start www-data 21802 0.0 0.2 331788 11768 ? S 21:30 0:00 /usr/sbin/apache2 -k start root 21849 0.0 0.0 16016 1116 pts/1 R+ 21:31 0:00 ps aux root 22363 0.0 0.0 0 0 ? S Oct25 0:09 [pdflush]
Do you have any recent fcgi processes die in your error.log? I wonder if the processes are spawning because they are waiting on mysql. I still have concern over your long queries being over 10 sec. It tells me that those are probably mysql writing a temp table out to disk...and disk tables for a busy server like yours are deadly. (still just a theory...but it would be interesting to see your disk access read/write during those maxcpu times.) This is related to if your website has a lot of fulltext searches. Are these custom queries or a packaged script? I would also increase your key_buffer_size to at least 128M and see how it goes. I keep mine the same rough size as my index files. You can reduce it once you solve some problems.
