Description
I just pull and ran 3.9.0 container today, and let celery use it for task queue. for every celery task sent out, the celery worker always report a error message it's disconnected from rabbitmq server, then it recover from it. I tried run 3.9.0 and 3.9.0-alpine, same result, and tried on two different server, one centos 7 , another centos8, same error.
however by switch to 3.8.9 image, everything is fine.
here it's linux dmesg output, my guess is somehow the 3.9.0 version cause docker's veth interface temporary down .
[Wed Jul 28 15:27:00 2021] docker0: port 1(vetha2ca9f3) entered blocking state
[Wed Jul 28 15:27:00 2021] docker0: port 1(vetha2ca9f3) entered disabled state
[Wed Jul 28 15:27:00 2021] device vetha2ca9f3 entered promiscuous mode
[Wed Jul 28 15:27:00 2021] IPv6: ADDRCONF(NETDEV_UP): vetha2ca9f3: link is not ready
[Wed Jul 28 15:27:00 2021] eth0: renamed from veth5dd6688
[Wed Jul 28 15:27:00 2021] IPv6: ADDRCONF(NETDEV_UP): eth0: link is not ready
[Wed Jul 28 15:27:00 2021] IPv6: ADDRCONF(NETDEV_UP): eth0: link is not ready
[Wed Jul 28 15:27:00 2021] IPv6: ADDRCONF(NETDEV_CHANGE): eth0: link becomes ready
[Wed Jul 28 15:27:00 2021] IPv6: ADDRCONF(NETDEV_CHANGE): vetha2ca9f3: link becomes ready
[Wed Jul 28 15:27:00 2021] docker0: port 1(vetha2ca9f3) entered blocking state
[Wed Jul 28 15:27:00 2021] docker0: port 1(vetha2ca9f3) entered forwarding state
[Wed Jul 28 15:27:00 2021] userif-3: sent link down event.
[Wed Jul 28 15:27:00 2021] userif-3: sent link up event.