OP Cache大比拼: APC vs eAccelerator vs xcache, eAccelerator胜出
8月 25th, 2008 Posted in PHP < by Johnny Woo >
APC 3.0.19 (Alternative PHP Cache)
配置
extension=apc.so
apc.enabled=1
apc.shm_segments=1
apc.shm_size=128
apc.ttl=7200
apc.user_ttl=7200
apc.num_files_hint=1024
apc.mmap_file_mask=/tmp/apc.XXXXXX
apc.enable_cli=1
apc.enabled=1
apc.shm_segments=1
apc.shm_size=128
apc.ttl=7200
apc.user_ttl=7200
apc.num_files_hint=1024
apc.mmap_file_mask=/tmp/apc.XXXXXX
apc.enable_cli=1
结果
7303 fetches, 50 max parallel, 1.67458e+07 bytes, in 300 seconds
2293 mean bytes/connection
24.3433 fetches/sec, 55819.3 bytes/sec
msecs/connect: 3.27583 mean, 2999.74 max, 0.184 min
msecs/first-response: 2043.52 mean, 7853.21 max, 325.513 min
HTTP response codes:
code 200 -- 7303
2293 mean bytes/connection
24.3433 fetches/sec, 55819.3 bytes/sec
msecs/connect: 3.27583 mean, 2999.74 max, 0.184 min
msecs/first-response: 2043.52 mean, 7853.21 max, 325.513 min
HTTP response codes:
code 200 -- 7303
eAccelerator 0.9.5.3
配置
[eaccelerator]
extension=eaccelerator.so
eaccelerator.shm_size="16"
eaccelerator.cache_dir="/tmp/eaccelerator"
eaccelerator.enable="1"
eaccelerator.optimizer="1"
eaccelerator.check_mtime="1"
eaccelerator.debug="0"
eaccelerator.log_file = "/var/log/httpd/eaccelerator_log"
eaccelerator.filter=""
eaccelerator.shm_max="0"
eaccelerator.shm_ttl="0"
eaccelerator.shm_prune_period="0"
eaccelerator.shm_only="0"
eaccelerator.compress="1"
eaccelerator.compress_level="9"
extension=eaccelerator.so
eaccelerator.shm_size="16"
eaccelerator.cache_dir="/tmp/eaccelerator"
eaccelerator.enable="1"
eaccelerator.optimizer="1"
eaccelerator.check_mtime="1"
eaccelerator.debug="0"
eaccelerator.log_file = "/var/log/httpd/eaccelerator_log"
eaccelerator.filter=""
eaccelerator.shm_max="0"
eaccelerator.shm_ttl="0"
eaccelerator.shm_prune_period="0"
eaccelerator.shm_only="0"
eaccelerator.compress="1"
eaccelerator.compress_level="9"
结果
7568 fetches, 50 max parallel, 1.73534e+07 bytes, in 300.001 seconds
2293 mean bytes/connection
25.2266 fetches/sec, 57844.6 bytes/sec
msecs/connect: 0.39873 mean, 1.01 max, 0.173 min
msecs/first-response: 1972.44 mean, 5520.1 max, 635.565 min
HTTP response codes:
code 200 -- 7568
2293 mean bytes/connection
25.2266 fetches/sec, 57844.6 bytes/sec
msecs/connect: 0.39873 mean, 1.01 max, 0.173 min
msecs/first-response: 1972.44 mean, 5520.1 max, 635.565 min
HTTP response codes:
code 200 -- 7568
XCache v1.2.2
配置
extension="xcache.so"
[xcache.admin]
; Change xcache.admin.user to your preferred login name
xcache.admin.user = "admin"
; Change xcache.admin.pass to the MD5 fingerprint of your password
; Use md5 -s "your_secret_password" to find the fingerprint
xcache.admin.pass = "e10adc3949ba59abbe56e057f20f883e"
[xcache]
; Change xcache.size to tune the size of the opcode cache
xcache.size = 24M
xcache.shm_scheme = "mmap"
xcache.count = 2
xcache.slots = 8K
xcache.ttl = 0
xcache.gc_interval = 0
; Change xcache.var_size to adjust the size of variable cache
xcache.var_size = 8M
xcache.var_count = 1
xcache.var_slots = 8K
xcache.var_ttl = 0
xcache.var_maxttl = 0
xcache.var_gc_interval = 300
xcache.test = Off
xcache.readonly_protection = On
xcache.mmap_path = "/tmp/xcache"
xcache.coredump_directory = ""
xcache.cacher = On
xcache.stat = On
xcache.optimizer = On
[xcache.coverager]
xcache.coverager = Off
xcache.coveragedump_directory = ""
[xcache.admin]
; Change xcache.admin.user to your preferred login name
xcache.admin.user = "admin"
; Change xcache.admin.pass to the MD5 fingerprint of your password
; Use md5 -s "your_secret_password" to find the fingerprint
xcache.admin.pass = "e10adc3949ba59abbe56e057f20f883e"
[xcache]
; Change xcache.size to tune the size of the opcode cache
xcache.size = 24M
xcache.shm_scheme = "mmap"
xcache.count = 2
xcache.slots = 8K
xcache.ttl = 0
xcache.gc_interval = 0
; Change xcache.var_size to adjust the size of variable cache
xcache.var_size = 8M
xcache.var_count = 1
xcache.var_slots = 8K
xcache.var_ttl = 0
xcache.var_maxttl = 0
xcache.var_gc_interval = 300
xcache.test = Off
xcache.readonly_protection = On
xcache.mmap_path = "/tmp/xcache"
xcache.coredump_directory = ""
xcache.cacher = On
xcache.stat = On
xcache.optimizer = On
[xcache.coverager]
xcache.coverager = Off
xcache.coveragedump_directory = ""
结果
4539 fetches, 50 max parallel, 1.04079e+07 bytes, in 300 seconds
2293 mean bytes/connection
15.13 fetches/sec, 34693.1 bytes/sec
msecs/connect: 0.405883 mean, 0.999 max, 0.182 min
msecs/first-response: 3214.64 mean, 18691.3 max, 772.952 min
HTTP response codes:
code 200 -- 4539
2293 mean bytes/connection
15.13 fetches/sec, 34693.1 bytes/sec
msecs/connect: 0.405883 mean, 0.999 max, 0.182 min
msecs/first-response: 3214.64 mean, 18691.3 max, 772.952 min
HTTP response codes:
code 200 -- 4539
No OP-Cache
结果
1798 fetches, 50 max parallel, 3.9867e+06 bytes, in 300 seconds
2217.3 mean bytes/connection
5.99332 fetches/sec, 13289 bytes/sec
msecs/connect: 12.0802 mean, 2999.48 max, 0.203 min
msecs/first-response: 6966.64 mean, 59754 max, 381.002 min
41 timeouts
61 bad byte counts
HTTP response codes:
code 200 -- 1757
2217.3 mean bytes/connection
5.99332 fetches/sec, 13289 bytes/sec
msecs/connect: 12.0802 mean, 2999.48 max, 0.203 min
msecs/first-response: 6966.64 mean, 59754 max, 381.002 min
41 timeouts
61 bad byte counts
HTTP response codes:
code 200 -- 1757
结论:
eAccelerator 0.9.5.3
25.2266 fetches/sec
APC 3.0.19
24.3433 fetches/sec
XCache v1.2.2
15.13 fetches/sec
No OP-Cache
5.99332 fetches/sec
eA的效果最好.APC紧随其后.两者的性能表现基本相同.xcache最差.
8月 27th, 2008 at 1:05 上午
用的是Apache吗？也许用lighttpd对xcache公平一些。
博主，鄙人的blog现在搬家到了：http://www.csphp.com/ 原来的 csask站长日志 已经不使用了：）
8月 28th, 2008 at 4:24 下午
好的.我更改下好友链接
用的是apache 2.2.4的
因为这些都是PHP的插件.所以用lighttpd.我猜测可能不会有什么改变.
8月 31st, 2008 at 5:39 上午
[...] OP Cache大比拼: APC vs eAccelerator vs xcache, eAccelerator胜出 | 架构研究室 (tags: cache) [...]
11月 11th, 2008 at 3:09 下午
I have problem with these comparisons:
1) shm sizes are different:
APC: 128M
eA: 16M
XCache: 24M
2) Optimization turned for eA and XCache are turned on; this will impact performance quite a bit.
11月 11th, 2008 at 3:18 下午
内存方面我在最后整理的时候注意到了
但是有两点考虑让我最后决定忽略它
1.这次测试只是一个页面.所以OPCODE缓冲使用的空间16M也足够了
2.测试结果显示尽管XCache比eA拥有更多内存,但是测试结果却更差.说明内存不是一个,至少不是一个关键的影响因素
关于优化选项,eA和XCache都打开了,而APC似乎没有什么优化选项,或者是我忽略了.不过我曾测试过ea和xc各自打开和关闭优化选项所带来的表现,结果是在这个测试用例中,是可以忽略其影响的.
谢谢你的提问,这使得今后我们的测试会更严谨.