You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
This program is one that is supposed to run on demand and finishes in less than a second so I cannot increase startsecs. Let's say this program is failing to start and returns an exit code of 255. I see this in the supervisor log:
2014-03-20 12:05:12,836 INFO exited: myprog (exit status 255; not expected)
Even though the program dies in an unexpected way, supervisor tells my event listeners that the program went into an EXITED state rather than a FATAL state. If I were to change the configuration only by setting "autorestart = unexpected" then supervisor will correctly determine that the program died in an unexpected way and does try to restart it. But the program will never go into a FAILED state and supervisor will simply try to restart it forever, even in the face of a "startretries" setting. That's the bug: supervisor fails to put a program into a failed state under this configuration even when it recognizes that the program is failing.
This is with supervisor tag 3.0.
The text was updated successfully, but these errors were encountered:
Say I have a program configured like this:
This program is one that is supposed to run on demand and finishes in less than a second so I cannot increase startsecs. Let's say this program is failing to start and returns an exit code of 255. I see this in the supervisor log:
Even though the program dies in an unexpected way, supervisor tells my event listeners that the program went into an EXITED state rather than a FATAL state. If I were to change the configuration only by setting "autorestart = unexpected" then supervisor will correctly determine that the program died in an unexpected way and does try to restart it. But the program will never go into a FAILED state and supervisor will simply try to restart it forever, even in the face of a "startretries" setting. That's the bug: supervisor fails to put a program into a failed state under this configuration even when it recognizes that the program is failing.
This is with supervisor tag 3.0.
The text was updated successfully, but these errors were encountered: