Describe the bug
while streaming data to a pipeline job, stateful block gets stopped after several "ZMQListener poll timeout" warnings from wrapper.
After stop block still gets msgs from bus (but cannot process them correctly) and pod does not exit.
Expected behavior
if gets "ZMQListener poll timeout" which leads to stop also exit
To Reproduce
Steps to reproduce the behavior:
streaming pipeline
stateful blocks
run data through pipeline
Happned due to intensive, up and down of stateless. Too many sockets are created.
A certain fix in wrapper created the problem, since wrapper 2.5 DEV10 wrapper, problem is fixed
Reported by Yael:
HKube micro-service worker / wrapper
Describe the bug while streaming data to a pipeline job, stateful block gets stopped after several "ZMQListener poll timeout" warnings from wrapper. After stop block still gets msgs from bus (but cannot process them correctly) and pod does not exit.
Expected behavior if gets "ZMQListener poll timeout" which leads to stop also exit
To Reproduce Steps to reproduce the behavior: streaming pipeline stateful blocks run data through pipeline