Background
Implement the manager responsible for handling reinitiating failed upload retries and node restart scenarios. These methods could be placed in the Block Upload manager class or a separate class.

We may be able to for each file in the BlockStreamConfig::blockFileDir directory, call BlockUploadManager::onBlockClosed. This would happen probably in ServicesMain prior to starting to handle transactions.

See https://github.com/hashgraph/hedera-services/blob/16770-block-stream-bucket-uploader-design/hedera-node/docs/proposals/block-stream-bucket-uploader.md#configuration-classes for more details.

Acceptance Criteria

• Implement the above core behaviors
• Create scan logic for identifying uploads which need to be reinitiated.
  • The block stream upload manager could notify the reupload manager (or keep track itself) of upload which have failed and need be retried yet again in the future.
  • Likewise if the node is restarted, it will handle trying to upload all blocks currently in the root block file directory. It should scan for these files prior to the BlockStreamManager producing more blocks after restart.
• Add proper error handling and logging
• Create unit tests

Dependencies
N/A