Cloudron makes it easy to run web apps like WordPress, Nextcloud, GitLab on your server. Find out more or install now.


Timeout in ESPOcrm, need to restart Webservice



  • Hi,
    our problem is that espocrm stops responding. We got a red Timeout message in the espocrm UI.
    Using supervisorctl restart web in the Terminal resolves the problem.

    I did not find any hints in the logs (cloudron and espo)


  • Staff

    @jensbee4 I published a new package now where we delete the apache pid file on a restart. I am not 100% sure if that is the source of the problem, but it's worth trying.

    If it still gets stuck, can you capture the output of supervisorctl status web and ps aux? Thanks!



  • While stuck:

    root@XXX693:/app/code/5.9.4# ps aux
    USER       PID %CPU %MEM    VSZ   RSS TTY      STAT START   TIME COMMAND
    root         1  0.0  0.1  64068  9672 pts/0    Ss+  Sep25   3:17 /usr/bin/python /usr/bin/supervisord --configuration /etc/supervisor/supervisord.conf --nodaemon -i EspoCRM
    www-data    49  0.0  0.2 554456 22940 pts/0    Sl   Sep25   0:09 /usr/bin/php /app/code/current/websocket.php
    root      1899  0.0  0.0  18508  2716 pts/1    Ss+  Sep30   0:00 /bin/bash
    root      1915  0.0  0.0  46628  5632 pts/0    S    Sep30   0:02 /usr/bin/python /usr/bin/pidproxy /var/run/apache2/apache2.pid /bin/bash -c APACHE_CONFDIR='' source /etc/apache2/envvars && rm
    root      1916  0.0  0.4 695032 34364 pts/0    S    Sep30   0:06 /usr/sbin/apache2 -DFOREGROUND
    root      1920  0.0  0.0   4672   700 pts/0    S    Sep30   0:00 /bin/cat
    root      1921  0.0  0.0   4672   712 pts/0    S    Sep30   0:00 /bin/cat
    www-data  2132  0.0  0.4 698144 35320 pts/0    S    06:52   0:00 /usr/sbin/apache2 -DFOREGROUND
    www-data  2133  0.0  0.5 782308 45880 pts/0    Sl   06:55   0:00 /usr/sbin/apache2 -DFOREGROUND
    www-data  2326  0.0  0.4 698200 38348 pts/0    S    07:13   0:00 /usr/sbin/apache2 -DFOREGROUND
    www-data  2338  0.0  0.4 698136 39396 pts/0    S    07:15   0:00 /usr/sbin/apache2 -DFOREGROUND
    www-data  2348  0.0  0.4 698196 39700 pts/0    S    07:17   0:00 /usr/sbin/apache2 -DFOREGROUND
    www-data  2592  0.1  0.5 780220 42248 pts/0    Sl   07:48   0:00 /usr/sbin/apache2 -DFOREGROUND
    root      2601  0.8  0.0  18508  3376 pts/2    Ss   07:56   0:00 /bin/bash
    root      2616  0.0  0.0  34404  2888 pts/2    R+   07:56   0:00 ps aux
    

    Next output

    root@XXXa693:/app/code/5.9.4# supervisorctl status web
    web                              RUNNING   pid 1915, uptime 1 day, 0:20:56
    

    Restarting Service:
    web: stopped
    web: started


  • Staff

    @jensbee4 Ah thanks. I think the issue is that apache gets "stuck". I have seen this happen when I run stress on the PHP apps as well. For a start, are you able to give the EspoCRM app more memory under Resources? How much memory have you given it right now? I think that will mitigate the issue.



  • @girish does it make sense to replace apache with nginx or lighttpd (many other options that are more modern, reliable, lightweight, faster and consume less memory)?


  • Staff

    Maybe, if we can't get apache to work. Usually upstream always recommends apache, so it's easier for us.



  • Memory Limit is set to 1GB ( CPU Shares : 75 %)
    Will go to 2GB. I will report when it happend again.

    Regards Jens



  • ESPO hangs again
    ps aux:

    USER       PID %CPU %MEM    VSZ   RSS TTY      STAT START   TIME COMMAND
    root         1  0.0  0.0  63948  7312 pts/0    Ss+  Oct02   2:10 /usr/bin/python /usr/bin/supervisord --configuration /etc/supervisor/supervisord.conf --nodaemon -i EspoCRM
    root        37  0.0  0.0  46620  5016 pts/0    S    Oct02   0:10 /usr/bin/python /usr/bin/pidproxy /var/run/apache2/apache2.pid /bin/bash -c APACHE_CONFDIR='' source /etc/apache2/envvars && rm -f ${A
    www-data    38  0.0  0.2 554456 23084 pts/0    Sl   Oct02   0:03 /usr/bin/php /app/code/current/websocket.php
    root        41  0.0  0.3 695028 24060 pts/0    S    Oct02   0:25 /usr/sbin/apache2 -DFOREGROUND
    root        45  0.0  0.0   4672   740 pts/0    S    Oct02   0:00 /bin/cat
    root        46  0.0  0.0   4672   688 pts/0    S    Oct02   0:01 /bin/cat
    root        69  0.0  0.0  18508  2304 pts/1    Ss+  Oct02   0:00 /bin/bash
    www-data  1065  0.0  0.4 698148 34536 pts/0    S    05:50   0:00 /usr/sbin/apache2 -DFOREGROUND
    www-data  1068  0.0  0.4 698156 34496 pts/0    S    06:02   0:00 /usr/sbin/apache2 -DFOREGROUND
    www-data  1078  0.0  0.3 697832 30292 pts/0    S    06:27   0:00 /usr/sbin/apache2 -DFOREGROUND
    www-data  1088  0.0  0.4 698148 32612 pts/0    S    06:40   0:00 /usr/sbin/apache2 -DFOREGROUND
    www-data  1091  0.2  0.4 698172 36008 pts/0    S    06:58   0:00 /usr/sbin/apache2 -DFOREGROUND
    www-data  1102  0.6  0.4 698144 35748 pts/0    S    07:00   0:00 /usr/sbin/apache2 -DFOREGROUND
    root      1103  0.0  0.0  18508  3404 pts/2    Ss   07:01   0:00 /bin/bash
    root      1123  0.0  0.0  34404  2908 pts/2    R+   07:02   0:00 ps aux
    

    supervisorctl status web
    web RUNNING pid 37, uptime 3 days, 22:12:03


  • Staff

    @jensbee4 Thanks. In the latest EspoCRM package (1.14.0), you can now go to File Manager and edit apache/mpm_preform.conf. What you can do is to bump the MaxRequestWorkers to say 20 or so and then restart the app.



  • @girish
    OK, will try to change the config.

    I got following:

    root@bcd346e4-22d8-49d4-93fc-89f39a77a693:/app/code/6.0.0# supervisorctl restart webnt size.
    web: stopped
    web: ERROR (spawn error)
    root@bcd346e4-22d8-49d4-93fc-89f39a77a693:/app/code/6.0.0# supervisorctl restart web
    web: stopped
    web: ERROR (spawn error)
    root@bcd346e4-22d8-49d4-93fc-89f39a77a693:/app/code/6.0.0# supervisorctl restart web
    web: stopped
    web: started
    root@bcd346e4-22d8-49d4-93fc-89f39a77a693:/app/code/6.0.0#
    

    Is the a log to see something about the spawn error?


  • Staff

    @jensbee4 said in Timeout in ESPOcrm, need to restart Webservice:

    Is the a log to see something about the spawn error?

    Yes, it's all redirected to stdout. So you can see it in the cloudron logs view.


Log in to reply