KAFKA-7658: Add KStream#toTable to the Streams DSL by highluck · Pull Request #7985 · apache/kafka

highluck · 2020-01-19T11:44:14Z

Add KStream#toTable to the Streams DSL

Committer Checklist (excluded from commit message)

Verify design and implementation
Verify test coverage and CI build status
Verify documentation (including upgrade notes)

mjsax

Thanks for the PR!

I think we need to add a test that verifies that a KTable is only materialized if we need to materialize it (eg, only if Materialized.as(name/storeSupplier) is used, or if the KTable is used in a join) -- the simplest way might be to verify if a StateStore is added to the processor via Topology#describe()?

highluck · 2020-01-20T17:03:43Z

@mjsax
thank you for review
I learned a lot from your review!

And update the code

thank you

mjsax

Thanks for the quick update. I think we need to be a little bit smarter about the materialization decision and add more test for it.

vvcephei · 2020-01-23T02:43:58Z

Thanks for pointing out the auto-materialization use case, @mjsax . I've been bitten by failing to get this right in a couple of recent features. A good, realistic, test seems to be just attempting to use the new operator as one join input.

mjsax

Thanks for adding more test -- left couple of comments. (Did only look at the tests.)

I think it would be good to add an aggregation test, too: stream.toTable().groupBy(...).count() to verify that it works... Not sure if we would need a KTable#filter() (or `mapValue) test? Maybe to verify that a filter/mapValue does not trigger a materialization?

I had one more thought (\cc @vvcephei): how should we handle the case if the KStream key was changes when toTable() is called? We can either insert a repartition topic automatically or we can disallow this operation (because we cannot guarantee the order or the update to the table). Personally, I think that inserting a repartition topic automatically should be fine. In any case, we also need a test for this case, too.

highluck · 2020-01-25T03:20:29Z

@mjsax
I agree with inserting a repartition topic automatically

highluck · 2020-01-28T11:13:42Z

@mjsax
I'm sorry, but can you please review it again?

I added a test case and added logic to repartition when calling toTable.

Thank you!!

mjsax

Thanks for adding more tests -- I think the DSL code is still not translated correctly into a Topology -- details comments on the tests.

highluck · 2020-01-29T15:23:28Z

@mjsax
I’m code update
I added a graphNode called StreamToTableNode.
And test update
,repartion logic update

By fixing the test, I could understand what you had said before.
Thank you
I learned a lot.

mjsax

Thanks updating the PR.

I reviewed the main code, too, this time. Couple of minor comments. Also some follow up comments on the tests (overall they look good now).

@mumrah -- This PR is basically ready and I would love to get it into 2.5 release -- when do you plan to cut the release branch? I don't want to cherry-pick a feature into the release branch hence I think there are two options:

we get the PR finalized before you cut the release branch
we merge the PR as-is, and do a follow up to address the open comments (cherry-picking a follow up seems ok, as the feature itself would be merged already).

Thoughts?

mjsax · 2020-01-29T18:35:53Z

It would be better to verify the whole TopologyDescription -- as you reuse this method for two different test, you could pass the "expectedTopology" String as parameter into this method.

However, reading the test code below, the case for table-table join, for which the right-hand table is updated to trigger a join result is missing (this case only makes sense for table-table join, as for stream-table join the right hand side table updates only update the rhs table, but never trigger a join result). Hence, I think it would be best to not share code but have two test and add the missing case for the table-table join.

mjsax · 2020-01-29T18:41:43Z

This method is only used once -- we should inline the code and remove the method.

mjsax · 2020-01-29T18:45:36Z

Why do we need this and cannot pass props as in the other tests?

mjsax · 2020-01-29T18:46:17Z

Why do we need to set wall-clock time?

mjsax · 2020-01-29T22:31:46Z

nit: avoid unnecessary this. prefix

mjsax · 2020-01-29T22:31:52Z

nit: avoid unnecessary this. prefix

mjsax · 2020-01-29T22:31:55Z

nit: avoid unnecessary this. prefix

mjsax · 2020-01-29T22:36:26Z

I think nextNode is a confusion name; better tableParentNode?

mjsax · 2020-01-29T22:37:43Z

nodes is not a good name -> subTopologySourceNodes is better.

mumrah · 2020-01-30T02:53:59Z

This PR is basically ready and I would love to get it into 2.5 release -- when do you plan to cut the release branch?

@mjsax the plan is to cut the branch tomorrow (1/30). Looks like @highluck has made some changes since your last review. Barring any issues, let's merge this tonight/tomorrow and open follow up PRs to address any outstanding minor issues.

mjsax

LGTM -- I will do a minor follow up PR the next days (there are some typos in the JavaDocs). Merging to meet the feature freeze deadline.

Thanks for the PR @highluck!

highluck · 2020-01-30T10:05:05Z

@mjsax
Thank you!

Conflicts and/or compiler errors due to the fact that we temporarily reverted the commit that removes Scala 2.11 support: * SslAdminIntegrationTest: keep using JAdminClient, take upstream changes otherwise. * ReassignPartitionsClusterTest: keep using JAdminClient, take upstream changes otherwise. * KafkaApis: use `asScala.foreach` instead of `forEach`. # By Ismael Juma (3) and others # Via GitHub * apache-github/trunk: (22 commits) KAFKA-9437; Make the Kafka Protocol Friendlier with L7 Proxies [KIP-559] (apache#7994) KAFKA-9375: Add names to all Connect threads (apache#7901) MINOR: Introduce 2.5-IV0 IBP (apache#8010) KAFKA-8503; Add default api timeout to AdminClient (KIP-533) (apache#8011) Add retries to release.py script (apache#8021) KAFKA-8162: IBM JDK Class not found error when handling SASL (apache#6524) MINOR: Add explicit result type in public defs/vals (apache#7993) KAFKA-9408: Use StandardCharsets.UTF-8 instead of "UTF-8" (apache#7940) KAFKA-9474: Adds 'float64' to the RPC protocol types (apache#8012) KAFKA-9360: Allow disabling MM2 heartbeat and checkpoint emissions (apache#7887) KAFKA-7658: Add KStream#toTable to the Streams DSL (apache#7985) KAFKA-9445: Allow adding changes to allow serving from a specific partition (apache#7984) KAFKA-9422: Track the set of topics a connector is using (KIP-558) (apache#8017) KAFKA-9040; Add --all option to config command (apache#7607) KAFKA-4203: Align broker default for max.message.bytes with Java producer default (apache#4154) KAFKA-9426: Use switch instead of chained if/else in OffsetsForLeaderEpochClient (apache#7959) KAFKA-9405: Use Map.computeIfAbsent where applicable (apache#7937) KAFKA-9026: Use automatic RPC generation in DescribeAcls (apache#7560) MINOR: Remove unused fields in StreamsMetricsImpl (apache#7992) KAFKA-9460: Enable only TLSv1.2 by default and disable other TLS protocol versions (KIP-553) (apache#7998) ...

Follow up to original PR #7985 for KIP-523 (adding `KStream#toTable()` operator) - improve JavaDocs - add more unit tests - fix bug for auto-repartitioning - some code cleanup Reviewers: High Lee <yello1109@daum.net>, John Roesler <john@confluent.io>

Follow up to original PR apache#7985 for KIP-523 (adding `KStream#toTable()` operator) - improve JavaDocs - add more unit tests - fix bug for auto-repartitioning - some code cleanup Reviewers: High Lee <yello1109@daum.net>, John Roesler <john@confluent.io>

highluck requested review from bbejeck, guozhangwang and mjsax and removed request for bbejeck January 19, 2020 11:44

mjsax reviewed Jan 19, 2020

View reviewed changes

mjsax added the streams label Jan 19, 2020

highluck force-pushed the kstream-ktable branch 2 times, most recently from 972f4fa to 7c83fc3 Compare January 20, 2020 16:59

highluck requested a review from mjsax January 20, 2020 17:02

mjsax reviewed Jan 22, 2020

View reviewed changes

highluck force-pushed the kstream-ktable branch 2 times, most recently from a0e83df to 0467a61 Compare January 23, 2020 07:56

highluck commented Jan 23, 2020

View reviewed changes

Comment thread streams/src/main/java/org/apache/kafka/streams/kstream/internals/KStreamImpl.java Outdated

highluck requested review from mjsax and vvcephei January 23, 2020 08:38

mjsax reviewed Jan 24, 2020

View reviewed changes

highluck force-pushed the kstream-ktable branch 3 times, most recently from 7356184 to 7908066 Compare January 25, 2020 03:18

highluck requested a review from mjsax January 25, 2020 03:20

highluck force-pushed the kstream-ktable branch 4 times, most recently from b4282c5 to 43c219b Compare January 28, 2020 10:58

mjsax reviewed Jan 28, 2020

View reviewed changes

highluck force-pushed the kstream-ktable branch from 43c219b to 395dfb7 Compare January 29, 2020 10:09

highluck requested a review from mjsax January 29, 2020 10:18

highluck force-pushed the kstream-ktable branch 3 times, most recently from f071490 to a1cab2e Compare January 29, 2020 15:23

mjsax reviewed Jan 29, 2020

View reviewed changes

highluck force-pushed the kstream-ktable branch from a1cab2e to 79ebe16 Compare January 30, 2020 02:23

KAFKA-7658: Add KStream#toTable to the Streams DSL

604a182

highluck force-pushed the kstream-ktable branch from 79ebe16 to 604a182 Compare January 30, 2020 06:31

mjsax approved these changes Jan 30, 2020

View reviewed changes

mjsax merged commit 6b86af3 into apache:trunk Jan 30, 2020

mjsax mentioned this pull request Feb 9, 2020

KAFKA-7658: Follow up to original PR #8027

Merged

mjsax added the kip Requires or implements a KIP label Jun 12, 2020

Conversation

highluck commented Jan 19, 2020

Committer Checklist (excluded from commit message)

Uh oh!

mjsax left a comment • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

highluck commented Jan 20, 2020

Uh oh!

mjsax left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

vvcephei commented Jan 23, 2020

Uh oh!

Uh oh!

mjsax left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

highluck commented Jan 25, 2020 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

highluck commented Jan 28, 2020

Uh oh!

mjsax left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

highluck commented Jan 29, 2020

Uh oh!

mjsax left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

mjsax left a comment •

edited

Loading

highluck commented Jan 25, 2020 •

edited

Loading