apache/hadoop · error · IOException

Incorrect block size

Error message

Incorrect block size

What it means

Error "Incorrect block size" thrown in apache/hadoop.

Source

Thrown at hadoop-hdfs-project/hadoop-hdfs/src/main/java/org/apache/hadoop/hdfs/server/datanode/BlockRecoveryWorker.java:266

        newBlock.setNumBytes(finalizedLength);
        break;
      case RBW:
      case RWR:
        long minLength = Long.MAX_VALUE;
        for(BlockRecord r : syncList) {
          ReplicaState rState = r.rInfo.getOriginalReplicaState();
          if(rState == bestState) {
            minLength = Math.min(minLength, r.rInfo.getNumBytes());
            participatingList.add(r);
          }
          LOG.debug("syncBlock replicaInfo: block={}, from datanode {}, receivedState={}, " +
              "receivedLength={}, bestState={}", block, r.id, rState.name(),
              r.rInfo.getNumBytes(), bestState.name());
        }
        // recover() guarantees syncList will have at least one replica with RWR
        // or better state.
        if (minLength == Long.MAX_VALUE) {
          throw new IOException("Incorrect block size");
        }
        newBlock.setNumBytes(minLength);
        break;
      case RUR:
      case TEMPORARY:
        assert false : "bad replica state: " + bestState;
      default:
        break; // we have 'case' all enum values
      }
      if (isTruncateRecovery) {
        newBlock.setNumBytes(rBlock.getNewBlock().getNumBytes());
      }

      LOG.info("BlockRecoveryWorker: block={} (length={}), bestState={},"
              + " newBlock={} (length={}), participatingList={}",
          block, block.getNumBytes(), bestState.name(), newBlock,
          newBlock.getNumBytes(), participatingList);

View on GitHub (pinned to 2add963021)

Solutions

1. Verify the client computed the new block length correctly during recovery/append; retry the operation.
2. Check for version mismatch between client, NameNode, and DataNode causing disagreement on the block size.

When it happens

Trigger: During replica recovery synchronization, the target block size does not match the size agreed in the recovery protocol.

Common situations: A replica's on-disk size did not match the expected block size during recovery. Compare against other replicas and disk health before proceeding.

AI-assisted analysis of apache/hadoop@2add963021 (2026-08-22). Data as JSON: /api/errors/e873e320bf4a7167. Report an issue: GitHub.