feat: add the number of merge buffers used for druid emitter

[TessaIO]

Fixes #4345 .

Description

Add the current number of merge buffers used for historical, broker, and real-time.

Release note

---

Add the current number of merge buffers used for historical, broker, and real-time.

---

This PR has:

• been self-reviewed.
  • using the concurrency checklist (Remove this item if the PR doesn't have any relation to concurrency.)
• added documentation for new or modified features or behaviors.
• a release note entry in the PR description.
• added Javadocs for most classes and all non-trivial methods. Linked related entities via Javadoc links.
• added or updated version, license, or notice information in licenses.yaml
• added comments explaining the "why" and the intent of the code wherever would not be obvious for an unfamiliar reader.
• added unit tests or modified existing tests to cover new code paths, ensuring the threshold for code coverage is met.
• added integration tests.
• been tested in a test Druid cluster.

[TessaIO]

@gianm @LakshSingla can you PTAL?

[LakshSingla]

Thanks for the first contribution!

The code style would be failing on a few files. Checkout
a) https://github.com/apache/druid/blob/master/dev/intellij-setup.md#code-style to setup the editor to lint the code correctly
b) Running mvn checkstyle:checkstyle would help in confirming if the checkstyle would fail on the PR or not.

Regarding the PR:
a. I think there should be a query/merge/totalBuffers metric as well.
b. Apart from this, there can be a wait time associated with each query - the total time required to acquire the merge buffers - this should be a simple change given that now we acquire all the merge buffers in a single place
c. There should be tests for the changes made in the PR.

[LakshSingla]

This shouldn't be "allocated to broker". As per my perception, allocated to the broker means the total number of merge buffers present there. Also, there is no such thing as "groupBy merge query".

Suggested change

	|`query/merge/buffersUsed`|number of merge buffers allocated to broker while performing groupBy merge queries.|This metric is only available if the `QueryCountStatsMonitor` module is included.| |
	|`query/merge/buffersUsed`|Number of merge buffers used up to merge the results of group by queries.|This metric is only available if the `QueryCountStatsMonitor` module is included.| |

[LakshSingla]

We can probably have the correct implementation for this class, even if it is a test utility.

[TessaIO]

Thanks for the review @LakshSingla.
I addressed all of your comments. PTAL when you have time.

b. Apart from this, there can be a wait time associated with each query - the total time required to acquire the merge buffers - this should be a simple change given that now we acquire all the merge buffers in a single place

I didn't really get what needed to be done here. Can you elaborate more on this?

[LakshSingla]

@TessaIO Aplogy for getting back late on this PR. The PR LGTM, can you please address the test failures.

My original comment was related to adding more metrics around this area, such as the elapsed time between the query requesting the merge buffer, and the query getting allotted the merge buffer. That can be a part of subsequent PRs.