-
Bug
-
Resolution: Done-Errata
-
Undefined
-
rhel-8.6.0.z
-
None
-
grafana-container-9-43
-
None
-
Moderate
-
2
-
rhel-sst-pt-pcp
-
ssg_platform_tools
-
25
-
1
-
QE ack
-
False
-
-
None
-
PCP Sprint 1, PCP Sprint 2
-
Pass
-
RegressionOnly
-
-
x86_64
-
None
What were you trying to do that didn't work?
Deploying rhel8/grafana container image (https://catalog.redhat.com/software/containers/rhel8/grafana/5edf9c35dd19c7063a62aff6?architecture=amd64&image=6543c1717371c4bd30141cfc) within an OpenShift environment (via Grafana Operator, which populates the GF_INSTALL_PLUGINS environment variable) fails because the grafana-cli wrapper script expects the user running it to be grafana or root. In an OpenShift deployment, the UID doesn't match 1001.
sh-4.4$ whoami 1000700000 sh-4.4$ /usr/sbin/grafana-cli /usr/sbin/grafana-cli: please run this script as user "grafana" or root.
Please provide the package NVR for which bug is seen:
grafana-9.2.10-7.el8_9.x86_64
registry.redhat.io/rhel8/grafana:9 (also same issue with rhel8/grafana:7 and rhel9/grafana:9).
How reproducible:
Always
Steps to reproduce
- Deploy Grafana image with a UID other than 1001 in OpenShift and populate the GF_INSTALL_PLUGINS environment variable via Deployment manifest, with a plugin parameter (e.g. grafana-polystat-plugin).
Expected results
Ability to override the grafana-cli wrapper script so it will ignore the user not being UID 1001, removal of the wrapper script, or container image using /usr/libexec/grafana/grafana-cli directly (like is done in the upstream docker.io/grafana/grafana image, which works with Grafana Operator and GF_INSTALL_PLUGINS).
/usr/share/grafana $ whoami
1000700000
/usr/share/grafana $ grafana-cli
NAME:
Grafana CLI - A new cli application
[...]
Actual results
sh-4.4$ whoami 1000700000 sh-4.4$ /usr/sbin/grafana-cli /usr/sbin/grafana-cli: please run this script as user "grafana" or root.
- links to
-
RHBA-2024:127153 Updated rhel8/grafana container image