Killing an SCGI process panics Cherokee #577

Closed
Borkason opened this Issue Mar 24, 2013 · 2 comments

Comments

Projects
None yet
1 participant
Member

Borkason commented Mar 24, 2013

Original author: brian.gregor@gmail.com (October 09, 2009 18:03:26)

What steps will reproduce the problem?

  1. I am using Django. Cherokee launches Django as an SCGI information source:
    /usr/bin/python /dir/manage.py runfcgi method=threaded host=127.0.0.1
    port=3333 protocol=scgi errlog=/tmp/django_err.log
    outlog=/tmp/django_out.log pidfile=/tmp/django.pid
  2. Use a URL that calls into Django system. Kill the django process:
    'kill pid'
  3. Do #​2 (Github: #82) several times, eventually it will work

What is the expected output? What do you see instead?

What version of the product are you using? On what operating system?
Cherokee 0.99.24

Please provide any additional information below.

PID 25188: launched '/bin/sh -c exec /usr/bin/python /dir/manage.py runfcgi
method=threaded host=127.0.0.1 port=3333 protocol=scgi
errlog=/tmp/django_err.log outlog=/tmp/django_out.log
pidfile=/tmp/django.pid' with uid=5948, gid=100, env=inherited
Server is exiting..
PID 25184: exited re=0
PID 25171: exited re=0
PID 25172: exited re=0
PID 25173: exited re=0
PID 25174: exited re=0
PID 25175: exited re=0
PID 25176: exited re=0
PID 25177: exited re=0
PID 25178: exited re=0
PID 25179: exited re=0
PID 25180: exited re=0
PID 25181: exited re=0
PID 25182: exited re=0
PID 25183: exited re=0
PID 25185: exited re=0
PID 25186: exited re=0
PID 25187: exited re=0
PID 25189: exited re=0
PID 25188: exited re=0
Cherokee feels panic!

Original issue: http://code.google.com/p/cherokee/issues/detail?id=598

Member

Borkason commented Mar 24, 2013

From alobbs on October 17, 2009 12:49:53
rev. 3729 tries to fix this problem:

http://svn.cherokee-project.com/changeset/3729

Member

Borkason commented Mar 24, 2013

From alobbs on November 02, 2009 14:19:58
This bug seems to be fixed now.
I've just check again by launching cherokee-admin and killing it's SCGI backend process:

while :; do killall -v python; sleep 1; done

After a ~5 minutes (killed 300 times) everything continued working just fine.
Brian, thanks for reporting!

@Borkason Borkason closed this Mar 24, 2013

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment