Actions
Bug #2395
closedVery high CPU usage lighttpd 1.4.28/29/30/31-devel
Status:
Invalid
Priority:
Normal
Category:
-
Target version:
-
ASK QUESTIONS IN Forums:
Description
This is really annoying:
Without cron scheduler enabled it looks like this:
top - 19:01:04 up 1:19, 4 users, load average: 15.75, 13.76, 12.51 Tasks: 387 total, 15 running, 303 sleeping, 0 stopped, 69 zombie Cpu(s): 40.0%us, 58.2%sy, 0.0%ni, 0.3%id, 0.4%wa, 0.1%hi, 0.8%si, 0.0%st Mem: 16371016k total, 16231396k used, 139620k free, 1502892k buffers Swap: 8388600k total, 0k used, 8388600k free, 8430204k cached PID USER PR NI VIRT RES SHR S %CPU %MEM TIME+ COMMAND 4347 mysql 15 0 4489m 1.7g 5860 S 93.7 10.6 20:49.26 mysqld 6464 apache 25 0 347m 206m 936 R 61.0 1.3 31:28.61 lighttpd 6462 apache 25 0 347m 195m 936 R 58.3 1.2 29:05.13 lighttpd 6463 apache 25 0 347m 206m 980 R 57.9 1.3 30:00.73 lighttpd 16169 agona2 25 0 98.5m 19m 5236 R 30.3 0.1 0:00.77 php-cgi 16135 admin 25 0 0 0 0 Z 27.6 0.0 0:00.70 php-cgi <defunct> 16150 mario271 24 0 0 0 0 Z 27.6 0.0 0:00.70 php-cgi <defunct> 31764 root 18 0 1188m 1.1g 4632 R 26.8 7.4 26:47.32 lphp.exe 16146 magdusie 24 0 363m 31m 12m R 23.2 0.2 0:00.59 php-cgi 16171 spokozio 25 0 0 0 0 Z 23.2 0.0 0:00.59 php-cgi <defunct> 16086 dazir 21 0 0 0 0 Z 22.0 0.0 0:00.70 php-cgi <defunct> 16153 rafaltra 25 0 0 0 0 R 21.7 0.0 0:00.55 php-cgi 16121 wojadann 25 0 0 0 0 Z 20.5 0.0 0:00.52 php-cgi <defunct> 16085 monia_ma 18 0 0 0 0 Z 20.1 0.0 0:00.70 php-cgi <defunct> 16106 dazir 16 0 370m 39m 13m S 18.5 0.2 0:00.60 php-cgi 16070 spokozio 25 0 0 0 0 Z 18.1 0.0 0:00.57 php-cgi <defunct> 16068 besthurt 16 0 0 0 0 Z 17.3 0.0 0:00.84 php-cgi <defunct> 16181 admin 25 0 360m 30m 13m R 16.5 0.2 0:00.42 php-cgi 16199 dazir 25 0 361m 32m 12m S 16.1 0.2 0:00.41 php-cgi 16044 mojasilo 25 0 0 0 0 Z 14.6 0.0 0:00.69 php-cgi <defunct> 16152 fhu_sun 25 0 352m 22m 11m R 14.6 0.1 0:00.37 php-cgi 16071 admin 25 0 0 0 0 Z 13.4 0.0 0:00.82 php-cgi <defunct> 16039 zootechn 17 0 0 0 0 Z 11.8 0.0 0:00.89 php-cgi <defunct> 16080 fhu_sun 24 0 0 0 0 Z 8.7 0.0 0:00.59 php-cgi <defunct> 16074 fhu_sun 25 0 0 0 0 Z 7.9 0.0 0:00.46 php-cgi <defunct> 16056 eline2 22 0 0 0 0 Z 6.7 0.0 0:00.79 php-cgi <defunct> 15996 admin 17 0 0 0 0 Z 6.3 0.0 0:01.09 php-cgi <defunct> 16201 root 17 0 27676 13m 4140 S 4.7 0.1 0:00.12 php 16211 spokozio 25 0 151m 4772 3408 R 4.7 0.0 0:00.12 php-cgi 16053 smerfetk 15 0 0 0 0 Z 3.5 0.0 0:00.71 php-cgi <defunct> 16054 dazir 23 0 0 0 0 Z 3.1 0.0 0:00.77 php-cgi <defunct> 16213 apache 25 0 8 4 0 R 3.1 0.0 0:00.08 lighttpd 7428 root 15 0 13016 1368 820 R 2.8 0.0 0:02.37 top 16016 admin 17 0 0 0 0 Z 2.0 0.0 0:00.86 php-cgi <defunct> 4774 varnish 18 0 12.3g 380m 81m S 0.8 2.4 2:16.94 varnishd 5908 memcache 15 0 143m 77m 404 S 0.8 0.5 0:13.70 memcached 2179 root 10 -5 0 0 0 S 0.4 0.0 0:02.41 md2_raid1 3535 named 24 0 340m 61m 2592 S 0.4 0.4 0:35.41 named 4569 lxlabs 15 0 66468 7472 1484 S 0.4 0.0 0:03.89 kloxo.httpd 7848 matiowip 15 0 116m 3300 2560 S 0.4 0.0 0:00.45 pure-ftpd
With cron jobs it's even worse:
top - 22:21:13 up 67 days, 10:28, 1 user, load average: 25.13, 23.18, 21.14 Tasks: 579 total, 48 running, 457 sleeping, 0 stopped, 74 zombie Cpu(s): 0.2%us, 80.3%sy, 19.2%ni, 0.0%id, 0.0%wa, 0.1%hi, 0.2%si, 0.0%st Mem: 16378256k total, 16216396k used, 161860k free, 541292k buffers Swap: 8388600k total, 735000k used, 7653600k free, 4689408k cached
Logs:
2012-03-06 19:24:05: (log.c.166) server started 2012-03-06 19:24:42: (mod_rewrite.c.380) execution error while matching: -8 2012-03-06 19:40:30: (mod_rewrite.c.380) execution error while matching: -8 2012-03-06 19:54:04: (mod_rewrite.c.380) execution error while matching: -8 2012-03-06 19:59:47: (mod_rewrite.c.380) execution error while matching: -8
My config:
server.max-worker = 3 server.event-handler = "linux-sysepoll" server.network-backend = "linux-sendfile" server.max-connections = 1000 server.stat-cache-engine = "simple" server.max-fds = 10000 server.max-keep-alive-idle = 0 #server.max-keep-alive-requests = 0 server.max-read-idle = 10 server.max-write-idle = 30 server.max-request-size = 40960 server.follow-symlink="enable"
CentOS 5.7 64-bit on Intel Core i7 2600 3.4 GHz + 16 GB DDR3 RAM
Updated by marek about 12 years ago
Kernel: 2.6.18-274.18.1.el5 #1 SMP Thu Feb 9 12:45:44 EST 2012 x86_64 x86_64 x86_64 GNU/Linux
Updated by darix about 12 years ago
- Priority changed from Urgent to Normal
- one of your rewrite rules is running amok.
- why is there more than one lighty process?
- your php processes seems to die aswell (defunct processes)
Updated by marek about 12 years ago
I've made a benchmark right now and take a look at results:
Varnish + Lighttpd: Requests per second: 4.53 [#/sec] (mean) ??? Lighttpd: Requests per second: 4.71 [#/sec] (mean) Apache: Requests per second: 229.80 [#/sec] (mean)
All were taken on the same machine with the same PHP (5.3.10).
Updated by marek about 12 years ago
darix wrote:
- why is there more than one lighty process?
I use server.max-worker = 3
But there seems to be even 4 processes:
apache 13416 0.5 0.5 179152 92056 ? S 23:18 0:00 /usr/sbin/lighttpd -f /etc/lighttpd/lighttpd.conf apache 13473 26.4 0.9 346972 153608 ? S 23:18 0:24 /usr/sbin/lighttpd -f /etc/lighttpd/lighttpd.conf apache 13474 27.4 0.8 346972 146600 ? S 23:18 0:25 /usr/sbin/lighttpd -f /etc/lighttpd/lighttpd.conf apache 13475 36.5 1.0 347104 164352 ? S 23:18 0:34 /usr/sbin/lighttpd -f /etc/lighttpd/lighttpd.conf
Updated by stbuehler about 12 years ago
- Status changed from New to Invalid
- Target version deleted (
1.4.31)
- I can't see any config
- No lighttpd version
- Broken php setup
- No support for server.max-worker != 1
- As darix pointed out, you probably have broken rewrite rules.
Actions
Also available in: Atom