ferenc-csaky commented on PR #5: URL: https://github.com/apache/flink-connector-hbase/pull/5#issuecomment-1354763682
I started to work on this, but I think there are things to discuss: * The HBase (and also Kafka) heavily relies on the custom JUnit4 based resources defined in `flink-end-to-end-tests-common` (e.g. `LocalStandaloneFlinkResource`, `DownloadCache`). On the short term the necessary calsses can be copied over to achieve the e2e test externalization ASAP and at least be able to execute the test, but on the longer term a common module probably would be useful. We can say every connector should take care about these things on their own, but IMO things like running flink, downloading artifacts will be required by multiple connector E2E tests, so eventually they will duplicate it. WDYT? * The HBase e2e test also relies on the `flink-sql-client-test` JAR, which is currently not deployed to maven. Would that be an option to deploy that as well for future releases so it is not necessary to put thogether the deps by hand? * The one and only HBase E2E test is [ignored](https://github.com/apache/flink/blob/b29d1055f9225d189603b5aeeacf3199e4c98636/flink-end-to-end-tests/flink-end-to-end-tests-hbase/src/test/java/org/apache/flink/tests/util/hbase/SQLClientHBaseITCase.java#L70) since 2021 April (according to Jira, that issue should be resolved but we never removed the `@Ignore` itself), so do we still consider this as a release blocker? It never ran during the whole 1.16 dev. cycle. WDYT? Maybe it would worth to discuss this in the dev channel in a more general manner, but I felt like posting it here first, cause I phrased everything from the HBase perspective, where these apply. cc @MartijnVisser @zentol -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: [email protected] For queries about this service, please contact Infrastructure at: [email protected]
