cni_ds_init should not be pid1 of cni daemon pod
Bug #1792539 reported by
Antoni Segura Puimedon
This bug affects 1 person
Affects | Status | Importance | Assigned to | Milestone | |
---|---|---|---|---|---|
kuryr-kubernetes |
Fix Released
|
Undecided
|
Antoni Segura Puimedon |
Bug Description
In containerized CNI, the executable that docker/crio runs is cni_ds_init. cni_ds_init is a shell script that sets up the environment and finally either runs the daemon executable or sleeps, depending of whether the deployment calls for a CNI daemon or just an executable.
The problem with doing that is that the PID 1 will be the shell script that is not doing anything once it reaches the cni daemon executable or sleep. This means the reported main process for the container/pod will not be
the one doing the job.
description: | updated |
To post a comment you must log in.
Fix proposed to branch: master /review. openstack. org/602573
Review: https:/