Command [/usr/local/bin/dub path /var/lib/kafka/data writable] FAILED !

I’m using kafka image v6.1.4-2-ubi8 pulled from https://hub.docker.com/layers/confluentinc/cp-kafka/6.1.4-2-ubi8/images/sha256-f3edca939e7484f5930eb91a8a42c5a30bf0b8e2556400c6677509b20cb26510?context=explore

Building my own Docker using above and using it in a kubernetes pod

While trying to run the image, I am getting the mentioned error

===> User
uid=1000670000(1000670000) gid=0(root) groups=0(root),1000670000
===> Configuring ...
SSL is enabled.
===> Running preflight checks ...
===> Check if /var/lib/kafka/data is writable ...
Command [/usr/local/bin/dub path /var/lib/kafka/data writable] FAILED !

I tried a lot of changes but nothing seems to work

USER root

RUN yum -y update --disableplugin=subscription-manager && \ 
	yum -y upgrade --disableplugin=subscription-manager && \
	groupadd -g 1001 fcigroup && \
  useradd -u 1001 -g 1001 -m fciuser && \
  chown -R fciuser: /var/lib/kafka && \
  chown -R fciuser: /home && \
  chown -R fciuser: /etc && \
  chown -R fciuser: /var/lib/kafka/data && \
  chmod -R 777 /var/log/kafka && \
  chmod -R 777 /home && \
  chmod -R 777 /etc/kafka && \
  chmod -R 777 /var/lib/kafka && \
  chmod -R 777 /var/lib/kafka/data

USER 1001

I even removed USER 1001 but even that didn’t seem to solve my issue

USER root

RUN yum -y update --disableplugin=subscription-manager && \ 
	yum -y upgrade --disableplugin=subscription-manager && \
  chmod -R a+rwx /var/log/kafka && \
  chmod -R a+rx /home && \
  chmod -R a+rwx /var/lib/kafka && \
  chmod -R a+rwx /var/lib/kafka/data && \
  chmod -R a+rwx /etc/kafka

I had to add /etc/kafka because earlier even that was throwing similar error

Any help is appreciated

[andrewegel]

yes, we are deploying on an openshift cluster… Is there an issue I can track or will you be using this one?

Great, thanks for confirming - We can track things here. I have some in-flight PRs I’ll list here that I believe will resolve things:

• https://github.com/confluentinc/kafka-images/pull/133
• https://github.com/confluentinc/schema-registry-images/pull/52
• https://github.com/confluentinc/kafka-rest-images/pull/33
• https://github.com/confluentinc/kafka-mqtt-images/pull/26
• https://github.com/confluentinc/control-center-images/pull/48

Internally we’re testing these changes to validate things are resolved before we merge & do mass rebuilds & re-releases of images. ETA is within the next couple of weeks. I will update things here as we make progress.

[andrewegel]

Hello folks-

After verifying these fixes worked in an OpenShift cluster with a random-uid-but-root-group runAs user, we have verified that this issue is no longer present. For the time being, I have pushed new a new 6.2.2-3-ub8 tag, and updated the 6.2.2 tag to these new images.

The following releases will also see this update something this week (I will update this issue as we release them):

• 5.4.6
• 5.5.7
• 6.0.5
• 6.1.4
• 7.0.1

[Edit]

• Pushed new tags 5.4.6, 5.4.6-3-deb8, 5.4.6-3-deb9, 5.4.6-3-ubi8 on 01-10-2022
• Pushed new tags 5.5.7, 5.5.7-3-deb8, 5.5.7-3-deb9, 5.5.7-3-ubi8 on 01-10-2022
• Pushed new tags 6.0.5, 6.0.5-3-ubi8 on 01-11-2022
• Pushed new tags 6.1.4, 6.1.4-3-ubi8 on 01-11-2022
• Pushed new tags 6.2.2, 6.2.2-3-ubi8 on 01-10-22
• Pushed new tags 7.0.1, 7.0.1-3-ubi8, latest, latest-ubi8 on 1-11-2022