日志
关于nginx性能优化CPU参数worker_cpu_affinity使用说明
|||
Nginx性能优化CPU参数worker_cpu_affinity使用说明
官方说明：http://wiki.nginx.org/NginxHttpMainModule#worker_cpu_affinit
官方说明：http://wiki.nginx.org/NginxHttpMainModule#worker_cpu_affinit
引用
worker_cpu_affinity
Syntax: worker_cpu_affinity cpumask [cpumask...]
Default: none
Linux only.
With this option you can bind the worker process to a CPU, it calls sched_setaffinity().
For example,
Bind each worker process to one CPU only.
Bind the first worker to CPU0/CPU2, bind the second worker to CPU1/CPU3. This is suitable for HTT.
Syntax: worker_cpu_affinity cpumask [cpumask...]
Default: none
Linux only.
With this option you can bind the worker process to a CPU, it calls sched_setaffinity().
For example,
worker_processes 4;
worker_cpu_affinity 0001 0010 0100 1000;
worker_cpu_affinity 0001 0010 0100 1000;
Bind each worker process to one CPU only.
worker_processes 2;
worker_cpu_affinity 0101 1010;
worker_cpu_affinity 0101 1010;
Bind the first worker to CPU0/CPU2, bind the second worker to CPU1/CPU3. This is suitable for HTT.
一个服务器为Q9300单cpu，4核心，于是可以直接用官方的配置：worker_cpu_affinity 0001 0010 0100 1000;
效果如下，看起来每个cpu核心都比较平均：
top - 21:17:56 up 4 days, 7:41, 3 users, load average: 0.25, 0.44, 0.47
Tasks: 238 total, 1 running, 237 sleeping, 0 stopped, 0 zombie
Cpu0 : 0.3%us, 0.3%sy, 0.0%ni, 94.2%id, 5.2%wa, 0.0%hi, 0.0%si, 0.0%st
Cpu1 : 0.3%us, 0.0%sy, 0.0%ni, 96.2%id, 0.6%wa, 0.6%hi, 2.2%si, 0.0%st
Cpu2 : 0.3%us, 0.0%sy, 0.0%ni, 93.5%id, 5.5%wa, 0.0%hi, 0.6%si, 0.0%st
Cpu3 : 0.0%us, 0.0%sy, 0.0%ni, 96.5%id, 3.5%wa, 0.0%hi, 0.0%si, 0.0%st
Mem: 8119372k total, 3641940k used, 4477432k free, 504900k buffers
Swap: 8385888k total, 10664k used, 8375224k free, 2407512k cached
PID USER PR NI VIRT RES SHR S %CPU %MEM TIME+ COMMAND
30743 apache 20 0 6632 3800 812 S 0.3 0.0 0:00.51 nginx
30745 apache 20 0 7156 4384 812 S 0.3 0.1 0:00.85 nginx
Tasks: 238 total, 1 running, 237 sleeping, 0 stopped, 0 zombie
Cpu0 : 0.3%us, 0.3%sy, 0.0%ni, 94.2%id, 5.2%wa, 0.0%hi, 0.0%si, 0.0%st
Cpu1 : 0.3%us, 0.0%sy, 0.0%ni, 96.2%id, 0.6%wa, 0.6%hi, 2.2%si, 0.0%st
Cpu2 : 0.3%us, 0.0%sy, 0.0%ni, 93.5%id, 5.5%wa, 0.0%hi, 0.6%si, 0.0%st
Cpu3 : 0.0%us, 0.0%sy, 0.0%ni, 96.5%id, 3.5%wa, 0.0%hi, 0.0%si, 0.0%st
Mem: 8119372k total, 3641940k used, 4477432k free, 504900k buffers
Swap: 8385888k total, 10664k used, 8375224k free, 2407512k cached
PID USER PR NI VIRT RES SHR S %CPU %MEM TIME+ COMMAND
30743 apache 20 0 6632 3800 812 S 0.3 0.0 0:00.51 nginx
30745 apache 20 0 7156 4384 812 S 0.3 0.1 0:00.85 nginx
最关键的地方没说清楚，怎样来表示每个CPU？
经过漫天的搜索和多次尝试发现
那么，16核的cpu每个进程分配到一个cpu就应该是
本站所在VPS的CPU为Intel(R) Xeon(R) CPU E5520 @ 2.27GHz x 2，因此写为：
worker_processes 2;
worker_cpu_affinity 01 10;
worker_cpu_affinity 01 10;
若要2核CPU开4进程，那么可以写为：
worker_processes 4;
worker_cpu_affinity 01 10 01 10;
worker_cpu_affinity 01 10 01 10;
2核是 01，四核是0001，8核是00000001，有多少个核，就有几位数，1表示该内核开启，0表示该内核关闭。
worker_processes最多开启8个，8个以上性能提升不会再提升了，而且稳定性变得更低，所以8个进程够用了。
配置完毕后，重启nginx ，执行/etc/init.d/nginx restart
测试nginx是否有用到多个CPU内核 ，在另一台机器上执行ab.exe -c 1000 -n 1000 http://你的网址
ab.exe是装apache后带的一个性能测试工具，它可以模拟多客户端的并发请求。
在服务器上执行top，然后按1，就可以看到CPU内核的工作情况。如果多个CPU内核的利用率都相差不多，证明nginx己经成功的利用了多核CPU。
测试结束后，CPU内核的负载应该都同时降低。
