rgw: reshard improvements

rgw: reshard improvements

Improve error log message when an expired reshard lock is renewed.

Add two new configurable options to manage resharding:
* rgw_reshard_batch_size : number of reshard entries to batch together
  before sending the operations to the CLS back-end.
* rgw_reshard_max_aio : maximum number of outstanding asynchronous i/o
  operations to allow at a time.

Alter rgw_reshard_bucket_lock duration default from 2 minutes to 6
minutes.

Add documentation, minimum values, tags, and service to a few rgw
reshard configuration options. Change some rgw_reshard_* options from
LEVEL_DEV to LEVEL_ADVANCED.

Signed-off-by: J. Eric Ivancich <ivancich@redhat.com>