apache/hadoop · error · IOException

Unexpected checksum mismatch while writing {} from {}

Error message

Unexpected checksum mismatch while writing {} from {}

What it means

Error "Unexpected checksum mismatch while writing {} from {}" thrown in apache/hadoop.

Source

Thrown at hadoop-hdfs-project/hadoop-hdfs/src/main/java/org/apache/hadoop/hdfs/server/datanode/BlockReceiver.java:517

      PacketHeader header = packetReceiver.getHeader();
      String specificOffset = "specific offsets are:"
          + " offsetInBlock = " + header.getOffsetInBlock()
          + " offsetInPacket = " + ce.getPos();
      LOG.warn("Checksum error in block "
          + block + " from " + inAddr
          + ", " + specificOffset, ce);
      // No need to report to namenode when client is writing.
      if (srcDataNode != null && isDatanode) {
        try {
          LOG.info("report corrupt " + block + " from datanode " +
                    srcDataNode + " to namenode");
          datanode.reportRemoteBadBlock(srcDataNode, block);
        } catch (IOException e) {
          LOG.warn("Failed to report bad " + block + 
                    " from datanode " + srcDataNode + " to namenode");
        }
      }
      throw new IOException("Unexpected checksum mismatch while writing "
          + block + " from " + inAddr);
    }
  }
  
    
  /**
   * Translate CRC chunks from the client's checksum implementation
   * to the disk checksum implementation.
   * 
   * This does not verify the original checksums, under the assumption
   * that they have already been validated.
   */
  private void translateChunks(ByteBuffer dataBuf, ByteBuffer checksumBuf) {
    diskChecksum.calculateChunkedSums(dataBuf, checksumBuf);
  }

  /** 
   * Check whether checksum needs to be verified.

View on GitHub (pinned to 2add963021)

Solutions

1. Investigate hardware issues (bad disk, faulty memory, NIC) on the sending or receiving host, since the checksum mismatch happened on data that was not expected to fail.
2. Enable dfs.datanode.drop.reason.behind to reduce spurious errors, and check preceding log lines for the upstream source of the corrupted bytes.

When it happens

Trigger: During block write, checksums computed over received packet data do not match the checksums transmitted with the packet, in a code path where a mismatch was not anticipated.

Common situations: Data written through the pipeline failed checksum verification on the receiving DataNode. Check disks, RAM, and NICs on the pipeline nodes; bad hardware is the usual cause.

AI-assisted analysis of apache/hadoop@2add963021 (2026-08-22). Data as JSON: /api/errors/70169a456262af02. Report an issue: GitHub.