Restart tunnel_packet_handler if it's not running in test_tunnel_memory_leak.py#5334
Restart tunnel_packet_handler if it's not running in test_tunnel_memory_leak.py#5334ZhaohuiS wants to merge 1 commit intosonic-net:masterfrom
Conversation
Signed-off-by: Zhaohui Sun <[email protected]>
|
Why |
The most of failed traceback loos like the following log, @theasianpianist Are you aware of it? How to resolve this issue? |
I agree, we should not need to manually start this service, it should be up automatically. I will investigate, let's hold off on this PR for now. |
|
Close this PR, the |
|
The issue was fixed in sonic-net/sonic-buildimage#10346 |
Description of PR
Summary:
Restart tunnel_packet_handler if it's not running in test_tunnel_memory_leak.py
Signed-off-by: Zhaohui Sun [email protected]
Type of change
Back port request
Approach
What is the motivation for this PR?
Sometimes, test case will fail due to tunnel_packet_handler is not running.
How did you do it?
Restart tunnel_packet_handler service in swss if it's not running at first time.
If it's still not running, the test case will fail. But first try restart will rescue tunnel_packet_handler back most of time.
How did you verify/test it?
run tests/dualtor/test_tunnel_memory_leak.py
Any platform specific information?
dualtor
Supported testbed topology if it's a new test case?
Documentation