Details
-
Bug
-
Status: Open
-
Major
-
Resolution: Unresolved
-
proton-c-0.37.0
-
None
-
None
-
None
-
centos 7
Description
I'm using the python wrapper and have seen this behaviour a number of times but don't know what precipitates it.
Symptoms are 100% cpu usage and a tcp connection status to 2 (send and receive) C qpid brokers in "close_wait" (the broker has timed out the connection due to a lack of heartbeat from the client)
It had, on startup, successfully connected to the broker but sometime later lost the connection. Scheduled tasks (_timers in _reactor.py) are not running (presumably starved by a tight loop in the process code)
On turning on debug logging it repeats (endlessly)
[DEBUG] <proton._reactor.Container object at 0x7fc91cf93310> Yielding
(presumably from line 251 of _reactor.py)