Derive aws sdk retry count from request count

[ljw9111]

Description

For retry count metric for Trino S3 file system, we are more interested in RetryCount than HttpClientRetryCount. But Trino was only tracking HttpClientRetryCount and not RetryCount.

Definition from aws java sdk v1 github is as follows
RetryCount - Number of retries of AWS SDK sending a request to AWS.
HttpClientRetryCount - Number of retries of the underlying http client library in sending a request to AWS.

And RetryCount count should be derived from RequestCount (i.e. RetryCount = RequestCount - 1) as in here

Additional context and related issues

Release notes

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

## Hive
* Fix reporting of AWS SDK client retry count metrics to report both client level logical retries and lower level HTTP client retries separately. Before the change, only tracking HttpClientRetryCount was collected and exposed via JMX as "awsretrycount" on the relevant JMX tables. Now, "awsretrycount" represents the logical retries and the HTTP client retries are reported as "awshttpclientretrycount" on tables such as, e.g.: "io.trino.hdfs.s3:name=hive,type=trinos3filesystem" or "trino.plugin.hive.metastore.glue:name=hive,type=gluehivemetastore". ({issue}`24606`)

[pettyjamesm]

The code change LGTM, can we catalog the JMX tables that are affected and include the column names in the release notes section on the PR?