Additional integration test checks

[arhimondr]

Description

• Ensure queries run by the AbstractTestQueryFramework have their final query info set and no tasks are left behind abandoned
• Limit query execution thread pool size to detect potential thread leaks (if a query scheduler gets abandoned upon query completion)

Non-technical explanation

N/A

Release notes

(X) This is not user-visible or docs only and no release notes are required.
( ) Release notes are required, please propose a release note for me.
( ) Release notes are required, with the following suggested text:

# Section
* Fix some things. ({issue}`issuenumber`)

[findepi]

Will it be not flaky as assertMemoryPoolReleased was?

Can you stress test on CI?

[arhimondr]

@findepi I introduced these checks when working on #14205 and have done many runs. I haven't noticed any tests being flaky due to these additional checks. But there's always a chance for a rare race condition.

[losipiuk]

there may be some tests which schedule queries in the background and do not necessarily wait for their completion. But we can address that if we see this becoming flaky.

[arhimondr]

There is one that I found: https://github.com/trinodb/trino/pull/14328/files#diff-c4d608eefc95a8f2dcee82dc30e0ee1f81313dded318f7fa0e381c1e54f4cd46R98.

[losipiuk]

nit: move to previous line

[arhimondr]

When I move it to the previous line it doesn't pass the style check

[losipiuk]

Interesting. I have checkstyle set up in IntelliJ and it did not complain. 🤷

[arhimondr]

Yeah, my Intellij doesn't complain either (though it is trying to format it with a new line when I try to autoformat). But then the check during the build fails.

[losipiuk]

nit: look unrelated - separate commit?

[arhimondr]

If the statement is not closed the query never finishes and doesn't receive a final TaskInfo making the newly introduced check fail

[losipiuk]

Still should be a preparatory commit. So adding checks and making project actually pass the checks are separate commits. (fixing makes sense even if we do not have checks).

[losipiuk]

CI: #14368