[ETCM-763] [ETCM-819] improve fast sync stability

[jvdp]

NB. this is based off of #977

This PR includes:

• improved (debug) logging in FastSync to help diagnose issues
• add (unique) names for every actor (for the benefit of logging)
• fix edge cases where Terminated message wasn't handled and FastSync was stopped
• fix bug where skeleton handler actor failure wouldn't be recovered from, causing fast sync to get stuck (currentSkeleton and skeletonHandler are now cleared.)
• fix issue(s) caused by stale state / actors hanging around after branch resolution (every child actor is now killed and a new SyncingHandler is instantiated.)

[leo-bogastry]

So one thing I have experienced when synching with Astor, was the fastSync failing without a proper error.
Here https://github.com/input-output-hk/mantis/pull/1001/files#diff-4b491aec4f7ba3bc8091f81d49100e0ccc38a2f396bcca38434118637e6c0befL637 when it calls validateBlocks, that method just returns Valid or Invalid.
So what do you think of also adding a log there?
This was my modification

def validateBlocks(
      requestedHashes: Seq[ByteString],
      blockBodies: Seq[BlockBody],
      log: LoggingAdapter
  ): BlockBodyValidationResult = {
    var result: BlockBodyValidationResult = Valid
    (requestedHashes zip blockBodies)
      .map { case (hash, body) => (blockchain.getBlockHeaderByHash(hash), body) }
      .forall {
        case (Some(header), body) =>
          val validationResult: Either[StdBlockValidator.BlockError, BlockValid] =
            validators.blockValidator.validateHeaderAndBody(header, body)
          result = validationResult.fold(
            error => {
              log.error("Block body validation failed with error {}", error)
              Invalid
            },
            _ => Valid
          )
          validationResult.isRight
        case _ =>
          result = DbError
          false
      }
    result
  }

[jvdp]

@LeonorLunatech thanks for the suggestion, I added the log message and also refactored it a bit.