[
https://issues.apache.org/jira/browse/FLINK-20118?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17235406#comment-17235406
]
Andrey Zagrebin edited comment on FLINK-20118 at 11/20/20, 7:20 AM:
--------------------------------------------------------------------
True, I looked into how TM failures could be done with MiniCluster. I only
found TestingMiniCluster::terminateTaskExecutor. If I try it, it currently
fails JVM for me abruptly. I can continue looking into this if there are no
better ideas
or did you mean some other approach for TM / JM failures?
On the other hand my own manual test is a bit broader (higher parallelism,
checks output order, standalone cluster). Having it as an e2e test does not
probably add much more than the existing IT case (if with TM / JM failures) but
it would also help to discover e.g. FLINK-20196.
Another idea (from offline discussion with [~trohrmann]) is to replace the
artificial source generator in the existing FileSink e2e test with the
FileSource. The e2e test for FileSink also tests failover. The ITCases are
indeed generally faster but we already have an e2e test for FileSink with
failover tests. Extending this e2e test should not explode test time too much.
Testing two things at once is generally not good but maybe it is ok for
Source/Sink pair and sparing test time. Individual aspects could be covered by
unit tests. Then we could remove ITCases and e2e test still covers a bit more
overall. Hence, this idea looks better to me atm.
What do you think about the last idea?
was (Author: azagrebin):
True, I looked into how TM failures could be done with MiniCluster. I only
found TestingMiniCluster::terminateTaskExecutor. If I try it, it currently
fails JVM for me abruptly. I can continue looking into this if there are no
better ideas
or did you mean some other approach for TM / JM failures?
On the other hand my own manual test is a bit broader (higher parallelism,
checks output order, standalone cluster). Having it as an e2e test does not
probably add much more than the existing IT case (if with TM / JM failures) but
it would also help to discover e.g. FLINK-20196.
Another idea (from offline discussion with [~trohrmann]) is to replace the
artificial source generator in the existing FileSink e2e test with the
FileSource. The ITCases are indeed generally faster but we already have an e2e
test for FileSink with failover tests. Extending this e2e test should not
explode test time too much. Testing two things at once is generally not good
but maybe it is ok for Source/Sink pair and sparing test time. Individual
aspects could be covered by unit tests. Then we could remove ITCases and e2e
test still covers a bit more overall. Hence, this idea looks better to me atm.
What do you think about the last idea?
> Test New File Source API - Continuous Streaming Execution
> ---------------------------------------------------------
>
> Key: FLINK-20118
> URL: https://issues.apache.org/jira/browse/FLINK-20118
> Project: Flink
> Issue Type: Sub-task
> Components: Connectors / Common
> Affects Versions: 1.12.0
> Reporter: Robert Metzger
> Assignee: Andrey Zagrebin
> Priority: Critical
> Fix For: 1.12.0
>
>
> ----
> [General Information about the Flink 1.12 release
> testing|https://cwiki.apache.org/confluence/display/FLINK/1.12+Release+-+Community+Testing]
> When testing a feature, consider the following aspects:
> - Is the documentation easy to understand
> - Are the error messages, log messages, APIs etc. easy to understand
> - Is the feature working as expected under normal conditions
> - Is the feature working / failing as expected with invalid input, induced
> errors etc.
> If you find a problem during testing, please file a ticket
> (Priority=Critical; Fix Version = 1.12.0), and link it in this testing ticket.
> During the testing keep us updated on tests conducted, or please write a
> short summary of all things you have tested in the end.
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
