频繁502错误，恳请军哥指教

yanqzh 发表于 2017-3-12 09:37:34

系统centos7 ，lnmp1.3，php7.0 ，mysql5.7.，内存4G，双核心。安装了memcached，启用了opcache
# netstat -anpo | grep "php-cgi" | wc -l
1

# free -m
         total    used    free shared buffers cached
Mem:       3832    3112    719       0    451    483
-/+ buffers/cache:    2177    1654
Swap:          0       0       0

#top
top - 09:10:35 up 23:57,1 user,load average: 0.32, 0.30, 0.27
Tasks: 167 total, 1 running, 166 sleeping, 0 stopped, 0 zombie
Cpu(s):0.3%us,0.2%sy,0.0%ni, 99.0%id,0.5%wa,0.0%hi,0.0%si,0.0%st
Mem: 3924164k total,3229268k used, 694896k free, 462044k buffers
Swap:    0k total,    0k used,    0k free, 497168k cached

PID USER    PRNIVIRTRESSHR S %CPU %MEM TIME+COMMAND
4805 www    20 0419m44m35m S1.31.2 0:03.47 php-fpm
4818 www    20 0419m49m37m S0.31.3 0:04.80 php-fpm
5909 root    20 0 15032 1268928 R0.30.0 0:00.05 top
30754 mysql 20 0 2295m 1.7g 7836 S0.3 45.6 7:01.13 mysqld
1 root    20 0 19356 1364 1048 S0.00.0 0:01.16 init

探针显示的系统负载均衡

nginx.conf部分配置：
worker_processes auto;
error_log/home/wwwlogs/nginx_error.logcrit;
pid    /usr/local/nginx/logs/nginx.pid;
worker_rlimit_nofile 51200;
worker_connections 51200;

   fastcgi_connect_timeout 300;
   fastcgi_send_timeout 300;
   fastcgi_read_timeout 300;
   fastcgi_buffer_size 64k;
   fastcgi_buffers 4 64k;
   fastcgi_busy_buffers_size 128k;
   fastcgi_temp_file_write_size 256k;
   gzip on;
   gzip_min_length1k;
   gzip_buffers 4 16k;
   gzip_http_version 1.1;
   gzip_comp_level 2;

php-fpm.conf部分配置：
listen = /tmp/php-cgi.sock
listen.allowed_clients = 127.0.0.1
pm.max_children = 200
pm.start_servers = 20
pm.min_spare_servers = 20
pm.max_spare_servers = 200
request_terminate_timeout = 100
request_slowlog_timeout = 0

opcache配置：

opcache.enable=1
zend_extension=/usr/local/php/lib/php/extensions/no-debug-non-zts-20151012/opcache.so
opcache.enable_cli=1
opcache.memory_consumption=128
opcache.interned_strings_buffer=8
opcache.max_accelerated_files=4000
opcache.revalidate_freq=60
opcache.fast_shutdown=1
opcache.file_cache=/tmp

探针里php相关参数

/usr/local/php/var/log/php-fpm.log部分错误：
NOTICE: fpm is running, pid 24916
NOTICE: ready to handle connections
WARNING: child 24920 exited on signal 11 (SIGSEGV - core dumped) after 30567.129989 seconds from start
NOTICE: child 2347 started
WARNING: child 24918 exited on signal 11 (SIGSEGV - core dumped) after 31322.164367 seconds from start
NOTICE: child 2608 started
WARNING: child 24922 exited on signal 11 (SIGSEGV - core dumped) after 38178.746541 seconds from start
NOTICE: child 4785 started
WARNING: child 24926 exited on signal 11 (SIGSEGV - core dumped) after 38180.449634 seconds from start
NOTICE: child 4787 started
WARNING: child 24925 exited on signal 11 (SIGSEGV - core dumped) after 38182.988441 seconds from start
NOTICE: child 4789 started
NOTICE: Finishing ...
NOTICE: exiting, bye-bye!
NOTICE: fpm is running, pid 4802
NOTICE: ready to handle connections

nginx_errors.log部分数据：
2017/03/11 17:37:40 21838#0: *1 connect() to unix:/tmp/php-cgi.sock failed (2: No such file or directory) while connecting to upstream, client: 106.38.241.141, server: aaaaaaa.com, request: "GET /index.php?title=Index.html&category-view-2-82_html=&mobileaction=toggle_view_desktop HTTP/1.1", upstream: "fastcgi://unix:/tmp/php-cgi.sock:", host: "aaaaaaaaaa.com"
2017/03/11 17:55:28 29396#0: *1 connect() to unix:/tmp/php-cgi.sock failed (2: No such file or directory) while connecting to upstream, client: 106.38.241.141, server: aaaaaaaa.com, request: "GET /title/%E5%A4%8F%E6%9E%AF%E8%8D%89 HTTP/1.1", upstream: "fastcgi://unix:/tmp/php-cgi.sock:", host: "aaaaaaaaa.com"

licess 发表于 2017-3-12 13:40:18

你这日子都不发同一时间段的，按nginx错误日子上是php-fpm挂了，按php-fpm日子是有程序执行超时，可以自己开慢日志看看具体的php程序

yanqzh 发表于 2017-3-13 09:26:19

回复 2# 的帖子

一、如何提高探针里的cache化内存呢？
对于cache化内存是否可以这样理解：
cache化的内存越大，服务器的cache功能就越好，网站响应及加载速度就越高？

那么问题来了，cache化内存如何提高？以前php5.X时，只要安装xcache，探针里cache化内存的展示图形长度，就必定比真实内存的展示图形长度要长，但是php7下，xcache无法安装。

二、php-fpm.conf内关于child的设置貌似无效，配置是：
pm.max_children = 200
pm.start_servers = 20
pm.min_spare_servers = 20
pm.max_spare_servers = 200
request_terminate_timeout = 600
request_slowlog_timeout = 600
slowlog = var/log/slow.log
但/usr/local/php/var/log/php-fpm.log提示：
WARNING: seems busy (you may need to increase pm.start_servers, or pm.min/max_spare_servers), spawning 8 children, there are 16 idle, and 30 total children
这是怎么回事呢？
# netstat -anpo | grep "php-cgi" | wc -l
显示数目是1

licess 发表于 2017-3-13 11:12:06

内存为系统自动管理，你就是用了xcache，没那么大的访问量，该占多少还是占多少

你把terminate 调小些

只有php 5.2上是php-cgi，5.3及之后早就都是php-fpm了

页: [1]

VPS侦探论坛's Archiver

频繁502错误，恳请军哥指教

回复 2# 的帖子