Gokul Kolady has posted comments on this change. ( 
http://gerrit.cloudera.org:8080/24368 )

Change subject: IMPALA-15028: Add Helm chart render assertions
......................................................................


Patch Set 48:

(2 comments)

http://gerrit.cloudera.org:8080/#/c/24368/48/helm/impala/tests/assert_chart.py
File helm/impala/tests/assert_chart.py:

http://gerrit.cloudera.org:8080/#/c/24368/48/helm/impala/tests/assert_chart.py@252
PS48, Line 252:         kudu_render.count("emptyDir: {}") >= 2,
This assertion can false-pass. Counting `emptyDir: {}` globally (`>= 2`) 
doesn’t guarantee both Kudu data volumes switched off PVCs; unrelated emptyDir 
blocks can satisfy the count.
Can we assert Kudu-specific behavior directly (e.g. verify both 
`kudu-master-data` and `kudu-tserver-data` render `emptyDir`, or assert both 
Kudu `claimName` entries are absent)?


http://gerrit.cloudera.org:8080/#/c/24368/48/helm/impala/tests/run-chart-tests.sh
File helm/impala/tests/run-chart-tests.sh:

http://gerrit.cloudera.org:8080/#/c/24368/48/helm/impala/tests/run-chart-tests.sh@103
PS48, Line 103: if helm template impala-secure-no-keytab "${CHART_DIR}" \
We validate missing keytab/TLS secret failures, but we don’t validate missing 
`security.kerberos.principal`.
Since this harness is meant to guard secure-cluster required inputs, can we add 
a negative test for `security.kerberos.enabled=true` without principal and 
assert the expected validation error?



--
To view, visit http://gerrit.cloudera.org:8080/24368
To unsubscribe, visit http://gerrit.cloudera.org:8080/settings

Gerrit-Project: Impala-ASF
Gerrit-Branch: master
Gerrit-MessageType: comment
Gerrit-Change-Id: I313be431e81a4b98370e51d298b7b2d35e0aded8
Gerrit-Change-Number: 24368
Gerrit-PatchSet: 48
Gerrit-Owner: Anubhav Jindal <[email protected]>
Gerrit-Reviewer: Abhishek Rawat <[email protected]>
Gerrit-Reviewer: Anubhav Jindal <[email protected]>
Gerrit-Reviewer: Gokul Kolady <[email protected]>
Gerrit-Reviewer: Impala Public Jenkins <[email protected]>
Gerrit-Reviewer: Jason Fehr <[email protected]>
Gerrit-Comment-Date: Tue, 28 Jul 2026 20:38:28 +0000
Gerrit-HasComments: Yes

Reply via email to