linux shell

关注公众号 jb51net

关闭
首页 > 脚本专栏 > linux shell > 启动supervisor报HTTP错误

普通用户启动supervisor报HTTP错误(strace)解决分析

作者:仁扬

这篇文章主要为大家介绍了普通用户启动supervisor报HTTP错误strace的问题解决分析,有需要的朋友可以借鉴参考下,希望能够有所帮助,祝大家多多进步,早日升职加薪

报错信息

公司的开发对生产环境都有普通用户 www 的权限,采用堡垒机登录到生产环境的机器。

默认 supervisor 使用 root 用户启动,开发没有权限直接修改配置和操作 supervisor 管理的进程,所以 supervisor 都采用 www 用户启动就解决问题了。

但是今天操作生产环境的时候,报的错误令人懵逼,SRE 同学折腾了好久,配置来配置去,搞不明白为啥 www 就是启动不了 supervisor !

报错信息是这样的:

[www@**************** ~]$ supervisord -c /etc/supervisord.conf
Error: Cannot open an HTTP server: socket.error reported errno.EACCES (13)
For help, use /usr/bin/supervisord -h

翻遍谷歌、百度,找到的资料,没有一个真的能解决了问题,焦头烂额。

突然想起 strace 这个命令,于是乎,我改成这样子执行:

[www@**************** ~]$ strace supervisord -c /etc/supervisord.conf

奇迹发生了!

read(6, "\0S(\1\0\0\0N(\3\0\0\0R'\0\0\0R \0\0\0R\24\0\0\0(\1\0\0"..., 4096) = 1381
read(6, "", 4096)                       = 0
close(6)                                = 0
close(5)                                = 0
close(4)                                = 0
getpid()                                = 3656
unlink("/var/run/supervisor/supervisor.sock.3656") = -1 ENOENT (No such file or directory)
socket(AF_UNIX, SOCK_STREAM, 0)         = 4
bind(4, {sa_family=AF_UNIX, sun_path="/var/run/supervisor/supervisor.sock.3656"}, 42) = -1 EACCES (Permission denied)
unlink("/var/run/supervisor/supervisor.sock.3656") = -1 ENOENT (No such file or directory)
write(2, "Error: Cannot open an HTTP serve"..., 75Error: Cannot open an HTTP server: socket.error reported errno.EACCES (13)
) = 75
write(2, "For help, use /usr/bin/superviso"..., 38For help, use /usr/bin/supervisord -h
) = 38
close(4)                                = 0
rt_sigaction(SIGINT, {sa_handler=SIG_DFL, sa_mask=[], sa_flags=SA_RESTORER, sa_restorer=0x7fa06ce997e0}, {sa_handler=0x7fa06d1b0750, sa_mask=[], sa_flags=SA_RESTORER, sa_restorer=0x7fa06ce997e0}, 8) = 0
close(3)                                = 0
close(11)                               = 0
exit_group(2)                           = ?
+++ exited with 2 +++

原来是没有权限,把 sock 文件挪到 www 用户下,问题解决。

关于 strace 命令

如果你的服务器还没安装 strace,可以用下面命令安装:

# Ubuntu / Debian
sudo apt install strace
# CentOS
yum install -y strace

如果一个进程已经在运行,可以通过 PID 跟踪程序的系统调用,使用 CTRL + C 中止跟踪:

[root@**************** ~]# strace -p 16701
strace: Process 16701 attached
restart_syscall(<... resuming interrupted read ...>) = 0
sendto(8, "*2\r\n$4\r\nLLEN\r\n$52\r\nhorizon:comma"..., 73, MSG_DONTWAIT, NULL, 0) = 73
poll([{fd=8, events=POLLIN|POLLERR|POLLHUP}], 1, 60000) = 1 ([{fd=8, revents=POLLIN}])
recvfrom(8, ":0\r\n", 8192, MSG_DONTWAIT, NULL, NULL) = 4
sendto(9, "*2\r\n$4\r\nLLEN\r\n$14\r\nqueues:defaul"..., 35, MSG_DONTWAIT, NULL, 0) = 35
poll([{fd=9, events=POLLIN|POLLERR|POLLHUP}], 1, 60000) = 1 ([{fd=9, revents=POLLIN}])
recvfrom(9, ":0\r\n", 8192, MSG_DONTWAIT, NULL, NULL) = 4
sendto(8, "*3\r\n$4\r\nHGET\r\n$21\r\nhorizon:queue"..., 55, MSG_DONTWAIT, NULL, 0) = 55
poll([{fd=8, events=POLLIN|POLLERR|POLLHUP}], 1, 60000) = 1 ([{fd=8, revents=POLLIN}])
recvfrom(8, "$18\r\n1.6690604487864007\r\n", 8192, MSG_DONTWAIT, NULL, NULL) = 25
wait4(16711, 0x7ffeaafa566c, WNOHANG|WSTOPPED, NULL) = 0
select(15, [12 14], [], [], {tv_sec=0, tv_usec=0}) = 0 (Timeout)
wait4(16712, 0x7ffeaafa566c, WNOHANG|WSTOPPED, NULL) = 0
select(17, [13 16], [], [], {tv_sec=0, tv_usec=0}) = 0 (Timeout)
wait4(16713, 0x7ffeaafa566c, WNOHANG|WSTOPPED, NULL) = 0
select(19, [15 18], [], [], {tv_sec=0, tv_usec=0}) = 0 (Timeout)
sendto(8, "*14\r\n$5\r\nHMSET\r\n$54\r\nhorizon:sup"..., 500, MSG_DONTWAIT, NULL, 0) = 500
sendto(8, "*4\r\n$4\r\nZADD\r\n$19\r\nhorizon:super"..., 99, MSG_DONTWAIT, NULL, 0) = 99
sendto(8, "*3\r\n$6\r\nEXPIRE\r\n$54\r\nhorizon:sup"..., 85, MSG_DONTWAIT, NULL, 0) = 85
poll([{fd=8, events=POLLIN|POLLERR|POLLHUP}], 1, 60000) = 1 ([{fd=8, revents=POLLIN}])
recvfrom(8, "+OK\r\n:0\r\n:1\r\n", 8192, MSG_DONTWAIT, NULL, NULL) = 13
sendto(8, "*3\r\n$5\r\nSETNX\r\n$29\r\nhorizon:moni"..., 58, MSG_DONTWAIT, NULL, 0) = 58
poll([{fd=8, events=POLLIN|POLLERR|POLLHUP}], 1, 60000) = 1 ([{fd=8, revents=POLLIN}])
recvfrom(8, ":0\r\n", 8192, MSG_DONTWAIT, NULL, NULL) = 4
rt_sigprocmask(SIG_BLOCK, [CHLD], [], 8) = 0
rt_sigaction(SIGCHLD, NULL, {sa_handler=SIG_DFL, sa_mask=[], sa_flags=0}, 8) = 0
rt_sigprocmask(SIG_SETMASK, [], NULL, 8) = 0
nanosleep({tv_sec=1, tv_nsec=0}, 0x7ffeaafa54a0) = 0
sendto(8, "*2\r\n$4\r\nLLEN\r\n$52\r\nhorizon:comma"..., 73, MSG_DONTWAIT, NULL, 0) = 73
poll([{fd=8, events=POLLIN|POLLERR|POLLHUP}], 1, 60000) = 1 ([{fd=8, revents=POLLIN}])
recvfrom(8, ":0\r\n", 8192, MSG_DONTWAIT, NULL, NULL) = 4
wait4(16711, 0x7ffeaafa566c, WNOHANG|WSTOPPED, NULL) = 0
select(15, [12 14], [], [], {tv_sec=0, tv_usec=0}) = 0 (Timeout)
wait4(16712, 0x7ffeaafa566c, WNOHANG|WSTOPPED, NULL) = 0
select(17, [13 16], [], [], {tv_sec=0, tv_usec=0}) = 0 (Timeout)
wait4(16713, 0x7ffeaafa566c, WNOHANG|WSTOPPED, NULL) = 0
select(19, [15 18], [], [], {tv_sec=0, tv_usec=0}) = 0 (Timeout)
sendto(8, "*14\r\n$5\r\nHMSET\r\n$54\r\nhorizon:sup"..., 500, MSG_DONTWAIT, NULL, 0) = 500
sendto(8, "*4\r\n$4\r\nZADD\r\n$19\r\nhorizon:super"..., 99, MSG_DONTWAIT, NULL, 0) = 99
sendto(8, "*3\r\n$6\r\nEXPIRE\r\n$54\r\nhorizon:sup"..., 85, MSG_DONTWAIT, NULL, 0) = 85
poll([{fd=8, events=POLLIN|POLLERR|POLLHUP}], 1, 60000) = 1 ([{fd=8, revents=POLLIN}])
recvfrom(8, "+OK\r\n:0\r\n:1\r\n", 8192, MSG_DONTWAIT, NULL, NULL) = 13
sendto(8, "*3\r\n$5\r\nSETNX\r\n$29\r\nhorizon:moni"..., 58, MSG_DONTWAIT, NULL, 0) = 58
poll([{fd=8, events=POLLIN|POLLERR|POLLHUP}], 1, 60000) = 1 ([{fd=8, revents=POLLIN}])
recvfrom(8, ":0\r\n", 8192, MSG_DONTWAIT, NULL, NULL) = 4
rt_sigprocmask(SIG_BLOCK, [CHLD], [], 8) = 0
rt_sigaction(SIGCHLD, NULL, {sa_handler=SIG_DFL, sa_mask=[], sa_flags=0}, 8) = 0
rt_sigprocmask(SIG_SETMASK, [], NULL, 8) = 0
nanosleep({tv_sec=1, tv_nsec=0}, ^Cstrace: Process 16701 detached
 <detached ...>

使用 -c 参数,可以得到各个系统调用的耗时、调用次数、错误数:

[root@**************** ~]# strace -p 16701 -c
strace: Process 16701 attached
^Cstrace: Process 16701 detached
% time     seconds  usecs/call     calls    errors syscall
------ ----------- ----------- --------- --------- ----------------
 54.00    0.000216           5        39           sendto
 11.25    0.000045           1        25           recvfrom
  9.25    0.000037           1        25           poll
  8.25    0.000033           5         6           nanosleep
  6.50    0.000026           1        21           select
  5.75    0.000023           1        21           wait4
  3.50    0.000014           1        14           rt_sigprocmask
  1.50    0.000006           0         7           rt_sigaction
  0.00    0.000000           0         1           restart_syscall
------ ----------- ----------- --------- --------- ----------------
100.00    0.000400                   159           total

使用 -tt 参数,显示时间戳(微妙),-T 参数显示系统调用的耗时,合起来用效果是这样的:

[root@**************** ~]# strace -p 16701 -tt -T
strace: Process 16701 attached
23:05:15.019519 restart_syscall(<... resuming interrupted read ...>) = 0 <0.915045>
23:05:15.934770 sendto(8, "*2\r\n$4\r\nLLEN\r\n$52\r\nhorizon:comma"..., 73, MSG_DONTWAIT, NULL, 0) = 73 <0.000080>
23:05:15.934975 poll([{fd=8, events=POLLIN|POLLERR|POLLHUP}], 1, 60000) = 1 ([{fd=8, revents=POLLIN}]) <0.000022>
23:05:15.935065 recvfrom(8, ":0\r\n", 8192, MSG_DONTWAIT, NULL, NULL) = 4 <0.000022>
23:05:15.935215 wait4(16711, 0x7ffeaafa566c, WNOHANG|WSTOPPED, NULL) = 0 <0.000023>
23:05:15.935301 select(15, [12 14], [], [], {tv_sec=0, tv_usec=0}) = 0 (Timeout) <0.000026>
23:05:15.935390 wait4(16712, 0x7ffeaafa566c, WNOHANG|WSTOPPED, NULL) = 0 <0.000018>
23:05:15.935446 select(17, [13 16], [], [], {tv_sec=0, tv_usec=0}) = 0 (Timeout) <0.000018>
23:05:15.935503 wait4(16713, 0x7ffeaafa566c, WNOHANG|WSTOPPED, NULL) = 0 <0.000023>
23:05:15.935574 select(19, [15 18], [], [], {tv_sec=0, tv_usec=0}) = 0 (Timeout) <0.000023>
23:05:15.935731 sendto(8, "*14\r\n$5\r\nHMSET\r\n$54\r\nhorizon:sup"..., 500, MSG_DONTWAIT, NULL, 0) = 500 <0.000059>
23:05:15.935848 sendto(8, "*4\r\n$4\r\nZADD\r\n$19\r\nhorizon:super"..., 99, MSG_DONTWAIT, NULL, 0) = 99 <0.000046>
23:05:15.935937 sendto(8, "*3\r\n$6\r\nEXPIRE\r\n$54\r\nhorizon:sup"..., 85, MSG_DONTWAIT, NULL, 0) = 85 <0.000041>
23:05:15.936028 poll([{fd=8, events=POLLIN|POLLERR|POLLHUP}], 1, 60000) = 1 ([{fd=8, revents=POLLIN}]) <0.000019>
23:05:15.936088 recvfrom(8, "+OK\r\n:0\r\n:1\r\n", 8192, MSG_DONTWAIT, NULL, NULL) = 13 <0.000018>
23:05:15.936278 sendto(8, "*3\r\n$5\r\nSETNX\r\n$29\r\nhorizon:moni"..., 58, MSG_DONTWAIT, NULL, 0) = 58 <0.000057>
23:05:15.936385 poll([{fd=8, events=POLLIN|POLLERR|POLLHUP}], 1, 60000) = 1 ([{fd=8, revents=POLLIN}]) <0.000019>
23:05:15.936441 recvfrom(8, ":0\r\n", 8192, MSG_DONTWAIT, NULL, NULL) = 4 <0.000017>
23:05:15.936505 rt_sigprocmask(SIG_BLOCK, [CHLD], [], 8) = 0 <0.000017>
23:05:15.936564 rt_sigaction(SIGCHLD, NULL, {sa_handler=SIG_DFL, sa_mask=[], sa_flags=0}, 8) = 0 <0.000023>
23:05:15.936624 rt_sigprocmask(SIG_SETMASK, [], NULL, 8) = 0 <0.000017>
23:05:15.936672 nanosleep({tv_sec=1, tv_nsec=0}, ^Cstrace: Process 16701 detached
 <detached ...>

常用的参数也就这么几个,如果记不住,直接看帮助吧:)

以上就是普通用户启动 supervisor 报 HTTP 错误(strace)的详细内容,更多关于启动 supervisor 报 HTTP 错误的资料请关注脚本之家其它相关文章!

您可能感兴趣的文章:
阅读全文