Make XContentBuilder in AliasActions build is_write_index field

[upgle]

Alias 'is_write_index' feature was introduced in 6.4 version. but, toXContent method of AliasActions does not build for is_write_index field.

so is_write_index can not be set with IndicesAliasesRequest (RestHighLevelClient)

You can see it in the code below.

elasticsearch/server/src/main/java/org/elasticsearch/action/admin/indices/alias/IndicesAliasesRequest.java

Line 448 in 237650e

public XContentBuilder toXContent(XContentBuilder builder, Params params) throws IOException {

Sample of RestHighLevelClient

It does not set is_write_index despite specifying options with writeIndex chaining method.

IndicesAliasesRequest request = new IndicesAliasesRequest();
IndicesAliasesRequest.AliasActions aliasAction =
        new IndicesAliasesRequest.AliasActions(IndicesAliasesRequest.AliasActions.Type.ADD)
                .index(index)
                .alias(alias)
                .filter(filter)
                .writeIndex(isWriteIndex); // this chaining method does not working.

request.addAliasAction(aliasAction);
return client.indices().updateAliases(request, RequestOptions.DEFAULT);

Notes

• I added test for this PR

[elasticmachine]

Pinging @elastic/es-core-infra

[talevy]

thank you @upgle for contributing this fix and bringing it to our attention!

Rest High Level Client was missed here.

Your changes look good, but I think they are missing three things

• the IndicesClientIT changes would benefit from randomization
• The IndicesAliasesRequest's hashCode/toString/equals are also missing writeIndex
• Add test on the server-side for IndicesAliasesRequestTests, where toXContent is tested. This means updating RandomAliasActionsGenerator.java to randomly include writeIndex.

I've addressed these concerns in a commit to make it clearer: talevy@b72608d

thanks!

[upgle]

Hi @talevy, Thank you for your kind comments.
I've modified this PR with reference to the code you shared. Can I get a review again, please?

[talevy]

thank you for for the update, I will run our continuous integration test suite against this to double check we didn't miss anything

[talevy]

@elasticmachine, test this please

[upgle]

@talevy
Build is failed here https://elasticsearch-ci.elastic.co/job/elastic+elasticsearch+pull-request/1418/
I found version confilict engine exceptoin.. so I rebased it on top of master.

[2018-10-30T19:30:22,495][DEBUG][o.e.a.g.TransportGetAction] [node-3] null: failed to execute [get [test_1][test][1]: routing [null]]
org.elasticsearch.transport.RemoteTransportException: [node-1][127.0.0.1:40018][indices:data/read/get[s]]
Caused by: org.elasticsearch.index.engine.VersionConflictEngineException: [test][1]: version conflict, current version [2] is different than the one provided [1]
at org.elasticsearch.index.engine.Engine.getFromSearcher(Engine.java:618) ~[elasticsearch-7.0.0-alpha1-SNAPSHOT.jar:7.0.0-alpha1-SNAPSHOT]
at org.elasticsearch.index.engine.InternalEngine.get(InternalEngine.java:659) ~[elasticsearch-7.0.0-alpha1-SNAPSHOT.jar:7.0.0-alpha1-SNAPSHOT]

[upgle]

@talevy
And also found.. MixedClusterClientYamlTestSuiteIT.test failures

Tests with failures:
  - org.elasticsearch.backwards.MixedClusterClientYamlTestSuiteIT.test {p0=search/10_source_filtering/_source_include}
  - org.elasticsearch.backwards.MixedClusterClientYamlTestSuiteIT.test {p0=search/10_source_filtering/_source include and _source in body}
  - org.elasticsearch.backwards.MixedClusterClientYamlTestSuiteIT.test {p0=get_source/70_source_filtering/Source filtering}
  - org.elasticsearch.backwards.MixedClusterClientYamlTestSuiteIT.test {p0=get/71_source_filtering_with_types/Source filtering}
  - org.elasticsearch.backwards.MixedClusterClientYamlTestSuiteIT.test {p0=explain/20_source_filtering/Source filtering}

I think we need to retry integration test again. This seems to be caused by this issue. 6256c33

[talevy]

thanks for applying the rebase. I will rerun the tests

[talevy]

@elasticmachine, test this please

[upgle]

Hi @talevy, test is just passed 🎉

[talevy]

thank you @upgle! I'll go ahead and merge this into the appropriate branches