Whenever a uwsgi process dies, in the middle of a spooled task execution, it is respawned.
Since this happens before the task is marked as processed, and removed from the spooler (or re-spooled with a later execution time), the respawned uwsgi process proceeds with reading the file in the spooler once again, untill the crash causes are resolved.
This can potentially cause hundreds of invocations per minutes, flooding the disk with log files.
A mechanism to control the re-spawning should be used.
Whenever a uwsgi process dies, in the middle of a spooled task execution, it is respawned.
Since this happens before the task is marked as processed, and removed from the spooler (or re-spooled with a later execution time), the respawned uwsgi process proceeds with reading the file in the spooler once again, untill the crash causes are resolved.
This can potentially cause hundreds of invocations per minutes, flooding the disk with log files.
A mechanism to control the re-spawning should be used.