…covery hook (#54729)

In #53961 we defer the construction of Searchable Snapshot
Directory's BlobContainer and a BlobStoreIndexShardSnapshot
instances so that these objects are created when they are first
accessed, which we expect to be during the recovery process. At
the same time, assertions were added to ensure that the
construction is effectively executed under a generic or snapshot
thread.

Sadly, this is not always the case because there is always a
short window of time between the creation of the IndexShard and
the time the objects are created during the recovery process. It
is also possible that other components of Elasticsearch trigger
the creation of the blob container and snapshot.

We identified the following places:

- computing avg shard size of index shards in
    IndexService.createShard() (this is triggered when relocating
    a primary shard under the cluster state applier thread)
- computing indices stats in TransportIndicesStatsAction which
    calls indexShard.storeStats() which calls
    estimateSizeInBytes() (this is triggered by
    InternalClusterInfoService under the management thread pool)
- computing shard store metadata in
    IndexShard.snapshotStoreMetadata while calls
    failIfCorrupted(Directory) (this is triggered by
    TransportNodesListShardStoreMetadata, executed under the
    fetch_shard_store thread pool)
- TransportNodesListGatewayStartedShards should also use
    failIfCorrupted(Directory) at some point (triggered by the
    GatewayAllocator and executed under the fetch_shard_started
    thread pool)

This commit changes the way BlobContainer and a
BlobStoreIndexShardSnapshot instances are created so that it does
not happen on first access anymore but the objects are now
created using a specific loadSnapshot() method.

This method is explicitly called during the pre-recovery phase.

Until this method is called, the SearchableSnapshotDirectory acts
as if it was empty: the listAll() method returns an empty
array. Having this behavior allows the identified access points
to not fail and not trigger the snapshot loading before we
explicitly load it in the pre-recovery hook.