nginxのパフォーマンスを向上させてみる
前回のエントリーでパフォーマンスの測定ができるようになりました。
今回はnginxのパフォーマンス向上に努めます。
なお、サイトの事前調査では最大リクエスト数の処理数が70なので、限界までリクエストするようにします。
最初に設定前のファイルを見てみましょう。
nginx.conf
user www-data;
worker_processes 4;
pid /run/nginx.pid;
events {
worker_connections 1024;
# multi_accept on;
}
http {
##
# Basic Settings
##
sendfile on;
tcp_nopush on;
tcp_nodelay on;
# keepalive_timeout 10;
# client_header_timeout 10;
# client_body_timeout 10;
# reset_timedout_connection on;
send_timeout 10;
types_hash_max_size 2048;
# server_tokens off;
# server_names_hash_bucket_size 64;
# server_name_in_redirect off;
include /etc/nginx/mime.types;
default_type application/octet-stream;
##
# SSL Settings
##
ssl_protocols TLSv1 TLSv1.1 TLSv1.2; # Dropping SSLv3, ref: POODLE
ssl_prefer_server_ciphers on;
##
# Logging Settings
##
access_log /var/log/nginx/access.log;
error_log /var/log/nginx/error.log;
##
# Gzip Settings
##
gzip on;
gzip_disable "msie6";
# gzip_vary on;
# gzip_proxied any;
# gzip_comp_level 6;
# gzip_buffers 16 8k;
# gzip_http_version 1.1;
# gzip_types text/plain text/css application/json application/javascript text/xml application/xml application/xml+rss text/javascript;
##
# Virtual Host Configs
##
include /etc/nginx/conf.d/*.conf;
include /etc/nginx/sites-enabled/*;
}
worker数の設定
worker_processes 4;
結果
Concurrency Level: 69
Time taken for tests: 4.444 seconds
Complete requests: 69
Failed requests: 0
Total transferred: 2385261 bytes
HTML transferred: 2334546 bytes
Requests per second: 15.53 [#/sec] (mean)
Time per request: 4443.779 [ms] (mean)
Time per request: 64.403 [ms] (mean, across all concurrent requests)
Transfer rate: 524.18 [Kbytes/sec] received
Connection Times (ms)
min mean[+/-sd] median max
Connect: 20 32 5.5 32 39
Processing: 359 2457 1250.9 2577 4403
Waiting: 315 2410 1250.7 2529 4356
Total: 379 2489 1256.1 2608 4442
Percentage of the requests served within a certain time (ms)
50% 2597
66% 3101
75% 3579
80% 3835
90% 4261
95% 4327
98% 4442
99% 4442
100% 4442 (longest request)
私の初期設定では4になっていました。
こちらをautoに切り替えて、nginxに判断させます。
autoの場合は、自動でコア数に応じてくれます。
lscpuでコア数がわかります。
lscpu Architecture: x86_64 CPU op-mode(s): 32-bit, 64-bit Byte Order: Little Endian CPU(s): 2 On-line CPU(s) list: 0,1 Thread(s) per core: 1 Core(s) per socket: 1 Socket(s): 2 NUMA node(s): 1 Vendor ID: GenuineIntel CPU family: 6 Model: 45 Model name: Intel(R) Xeon(R) CPU E5-2640 0 @ 2.50GHz Stepping: 7 CPU MHz: 2499.998 BogoMIPS: 4999.99 Hypervisor vendor: KVM Virtualization type: full L1d cache: 32K L1i cache: 32K L2 cache: 4096K NUMA node0 CPU(s): 0,1
今回の場合は2ですね。
なぜ、4にしていたのだろうか・・・orz
結果
Concurrency Level: 69
Time taken for tests: 4.041 seconds
Complete requests: 69
Failed requests: 0
Total transferred: 2385261 bytes
HTML transferred: 2334546 bytes
Requests per second: 17.07 [#/sec] (mean)
Time per request: 4041.382 [ms] (mean)
Time per request: 58.571 [ms] (mean, across all concurrent requests)
Transfer rate: 576.38 [Kbytes/sec] received
Connection Times (ms)
min mean[+/-sd] median max
Connect: 21 33 5.7 33 46
Processing: 346 2222 1104.8 2259 3993
Waiting: 303 2175 1104.9 2215 3944
Total: 371 2255 1110.2 2292 4039
Percentage of the requests served within a certain time (ms)
50% 2287
66% 2794
75% 3232
80% 3401
90% 3784
95% 3933
98% 4039
99% 4039
100% 4039 (longest request)
比較
Time per request: 4443.779 [ms] (mean) Time per request: 64.403 [ms] (mean, across all concurrent requests)
Time per request: 4041.382 [ms] (mean) Time per request: 58.571 [ms] (mean, across all concurrent requests)
わずかながらですが、早くなりました。
mulit accept
同時に受け入れるかどうかでの判定です。 こちらもonにしてあげましょう。
multi_accept on;
結果
Concurrency Level: 69
Time taken for tests: 3.443 seconds
Complete requests: 69
Failed requests: 0
Total transferred: 2385261 bytes
HTML transferred: 2334546 bytes
Requests per second: 20.04 [#/sec] (mean)
Time per request: 3443.017 [ms] (mean)
Time per request: 49.899 [ms] (mean, across all concurrent requests)
Transfer rate: 676.55 [Kbytes/sec] received
Connection Times (ms)
min mean[+/-sd] median max
Connect: 22 33 5.8 36 39
Processing: 247 1815 979.8 1883 3402
Waiting: 200 1766 977.1 1836 3359
Total: 270 1848 985.3 1920 3441
Percentage of the requests served within a certain time (ms)
50% 1919
66% 2331
75% 2694
80% 2841
90% 3207
95% 3368
98% 3441
99% 3441
100% 3441 (longest request)
どんどん早くなっていく。
Time per request: 3443.017 [ms] (mean) Time per request: 49.899 [ms] (mean, across all concurrent requests)
こんな設定するだけでも、トータルのレスポンスの処理速度が1,000msも短縮できました。
わずかな設定でここまで差が出るとは・・・
ちゃんとサーバーの設定を見直すことは大事ですね。
