Hive Connector with Amazon S3 documentation updates

[dnanuti]

Description

Documentation updates following up changes on Hive connector with Amazon S3: fix for Select pushdown for uncompressed files, addition of JSON support to Amazon S3 Select and usage of S3 Select scan range requests.
Relevant PRs:

• Enable Select pushdown on uncompressed files #12633
• Add S3 Select pushdown for JSON files #13354
• Disable S3Select pushdown when query not filters data #13477
• Enabled hive splits for uncompressed CSV files with S3 Select pushdown #13754
• Enable hive splits for uncompressed JSON files with S3 Select pushdown #14040

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:

[arhimondr]

@jhlodin Could you please take a look?

[jhlodin]

Thanks for writing docs! I suggested edits for the paragraph and a recommended anchor link.

Are any changes needed to the "Is S3 Select a good fit for my workload" section above this content?

[jhlodin]

Suggested change

	For uncompressed files, Scan Range feature of S3 Select is used.
	An Amazon S3 Select scan range request runs across the specified byte range.
	This range is aligned with the internal Hive splits for the query fragments
	that get pushed down to Select. Changes in the Hive connector performance
	tuning configuration properties would be reflected here as well.
	For uncompressed files, S3 Select scans ranges of bytes in parallel. The scan range
	requests run across the byte ranges of the internal Hive splits for the query fragments
	pushed down to S3 Select. Changes to the Hive catalog's :ref:`performance tuning
	configuration properties <hive-performance-tuning-configuration>` are reflected
	here as well.

[jhlodin]

To make the anchor link work, please add
.. _hive-performance-tuning-configuration:
above line 734 of the hive connector page (/connector/hive.rst)

[dnanuti]

Thanks a lot! Updated 👍

[dnanuti]

Can't reply to the above comment. There are no changes needed for "Is S3 Select a good fit for my workload" section.

[jhlodin]

It looks like only the ref link was added, can you apply the rest of the suggested edits?

Should look like the following:

For uncompressed files, S3 Select scans ranges of bytes in parallel. The scan range
requests run across the byte ranges of the internal Hive splits for the query fragments
pushed down to S3 Select. Changes to the Hive catalog's :ref:`performance tuning
configuration properties <hive-performance-tuning-configuration>` are reflected
here as well.

[dnanuti]

This was rephrased a bit on our side as well. From a technical perspective, I think we should say Hive connector, not Hive catalog, as this is related to how the connector works.

Does this work for you?

For uncompressed files, S3 Select scans ranges of bytes in parallel. The scan range
requests run across the byte ranges of the internal Hive splits for the query fragments
pushed down to S3 Select. Changes in the Hive connector :ref:`performance tuning
configuration properties <hive-performance-tuning-configuration>` are likely to impact
S3 Select pushdown performance.

[jhlodin]

Yep, that makes sense to me! Once that change is in, LGTM

[jhlodin]

LGTM % one last minor suggestion

[jhlodin]

"they are retrieving" -> "they retrieve"