fix Optimization: Use Lucene's TwoPhaseIterator for quick filtering t…

[csywxr]

Using Lucene's Two Phase Query to submit and optimize the match query, the query has been tested and found to be problem free

[cla-checker-service]

❌ Author of the following commits did not sign a Contributor Agreement:
3601f39, 5a42a2e, 2fec87f

Please, read and sign the above mentioned agreement if you want to contribute to this project

[csywxr]

Using Lucene's Two Phase Query submission to optimize match queries

[csywxr]

Using Lucene's Two Phase Query submission to optimize match queries

[elasticsearchmachine]

Pinging @elastic/es-search-foundations (Team:Search Foundations)

[javanna]

Heya @csywxr , thanks for opening this PR!

Could you please share more context on what originated this, what problems it solves, and also why it was opened against the 7.17 branch instead of main?

Thanks!

[csywxr]

Hello Javanna
The 7.17 version is commonly used in our project, but the query efficiency is not as high as expected. Through analyzing the source code of Elasticsearch in 7.17, we found that we can use Lucense's TwoPhaseIterator to improve query efficiency

[csywxr]

Heya @csywxr , thanks for opening this PR!

Could you please share more context on what originated this, what problems it solves, and also why it was opened against the 7.17 branch instead of main?

Thanks!

Hello Javanna
The 7.17 version is commonly used in our project, but the query efficiency is not as high as expected. Through analyzing the source code of Elasticsearch in 7.17, we found that we can use Lucense's TwoPhaseIterator to improve query efficiency

[javanna]

Thanks for getting back to me. Do you have benchmarks results of the impact? Asking because we have no plans to do any further 7.17 releases, bearing any catastrophic bugs found, and this does not seem like it's the case.

We would be open to making the change against main, and perhaps backport it to 8.19. Yet we need to work on adding tests and better understanding the usecase and gain.

Correct me if I am wrong but I care to specify that this isn't a change to the match query of the query DSL, but to the named queries feature.

Thanks again!

[csywxr]

Thanks for getting back to me. Do you have benchmarks results of the impact? Asking because we have no plans to do any further 7.17 releases, bearing any catastrophic bugs found, and this does not seem like it's the case.

We would be open to making the change against main, and perhaps backport it to 8.19. Yet we need to work on adding tests and better understanding the usecase and gain.

Correct me if I am wrong but I care to specify that this isn't a change to the match query of the query DSL, but to the named queries feature.

Thanks again!

I aggree making the change against main