We have recently observed deadlocks occurring in roughly 1/50 to 1/100 deploys of containers. These will begin booting then hang endlessly.
Environment Information
Provide at least:
- Docker images (on amd64)
jruby:9.4.12.1-jdk8
jruby:9.4.12.1-jdk21
Attached below are two thread dumps showing the same deadlock. NOTE: many containers booted without any issue, but infrequently we get pods that boot and have this same deadlock.
231621_deadlock.tdump.txt
231314_deadlock.tdump.txt