1. Please take a little time for this simple survey! Thank you for participating!
    Dismiss Notice
  2. Dear Pleskians, please read this carefully! New attachments and other rules Thank you!
    Dismiss Notice
  3. Dear Pleskians, I really hope that you will share your opinion in this Special topic for chatter about Plesk in the Clouds. Thank you!
    Dismiss Notice

plesk 9.5.4: cronjob apache reload race condition (websrvmng?)

Discussion in 'Plesk 9.x for Linux Issues, Fixes, How-To' started by Internet AGS, Dec 12, 2011.

  1. Internet AGS

    Internet AGS Guest

    0
     
    Hello,

    on one of our plesk systems, a race condition between creation of virtualhost configuration files and apache2 reload happens every two or three days. It happens either around 04:10h am or around 10:20h am, and we're pretty sure that it's triggered by one of the plesk cronjobs.

    The system runs plesk '9.5.4 Debian 5.0 95101209.05' with microupdates until #5. Appart from this issue, it runs as expected.

    The race condition results in apache2 crashing. The relevant lines from apache2 error log are:

    Please note that the missing configuration file is from another virtualhost every time. When the problem is detected (apache2 not responding any more), and we manually start apache2 again, the virtualhost configuration file in question is available again. Thus the problem is unrelated to the issue described here.

    I guess that for some reason apache2 is being reloaded before the plesk cronjob finished with creating all virtualhost configuration files. The apache2 reload is invoked by websrvmng from plesk, but I'm not sure, which cronjob starts this process.

    Does anybody else experience a similar issue? What further steps are required to debug and isolate this problem?

    Regards,
    Jonas Meurer for Internet AG
     
  2. Internet AGS

    Internet AGS Guest

    0
     
    do you have any suggestions for us, how to futher debug the problem? It's important for us to find the source of this problem and fix it.

    Thanks in advance.
     
  3. Internet AGS

    Internet AGS Guest

    0
     
    websrvmng seems responsible

    Hello again,

    we tracked the problem down to websrvmng process which seems to restart apache2 at situations where it's not ready to do so yet. Unfortunately we don't know what to do next. Please give us advice.

    Regards,
    jonas
     
Loading...