[HADOOP-1134] Block level CRCs in HDFS

• Type: New Feature
• Status: Closed
• Priority: Major
• Resolution: Fixed
• Fix Version/s: 0.14.0
• Assignee: Raghu Angadi
• Reporter: Raghu Angadi
• Created: 19/Mar/07 22:27
• Updated: 02/May/13 02:29
• Resolved: 16/Jul/07 21:43

Description

Currently CRCs are handled at FileSystem level and are transparent to core HDFS. See recent improvement HADOOP-928 ( that can add checksums to a given filesystem ) regd more about it. Though this served us well there a few disadvantages :

1) This doubles namespace in HDFS ( or other filesystem implementations ). In many cases, it nearly doubles the number of blocks. Taking namenode out of CRCs would nearly double namespace performance both in terms of CPU and memory.

2) Since CRCs are transparent to HDFS, it can not actively detect corrupted blocks. With block level CRCs, Datanode can periodically verify the checksums and report corruptions to namnode such that name replicas can be created.

We propose to have CRCs maintained for all HDFS data in much the same way as in GFS. I will update the jira with detailed requirements and design. This will include same guarantees provided by current implementation and will include a upgrade of current data.

Issue Links

• depends upon: HADOOP-1597 Distributed upgrade status reporting and post upgrade features.
• depends upon: HADOOP-1286 Distributed cluster upgrade
• depends upon: HADOOP-1470 Rework FSInputChecker and FSOutputSummer to support checksum code sharing between ChecksumFileSystem and block level crc dfs
• incorporates: HADOOP-1471 seekToNewSource() might not work correctly with Checksum failures.
• incorporates: HADOOP-1478 The blockStream of DFSClient.FSInputStream should not be buffered
• is depended upon by: HADOOP-1564 Write unit tests to detect CRC corruption
• relates to: HADOOP-1084 updating a hdfs file, doesn't cause the distributed file cache to update itself
• relates to: HADOOP-1086 spurious data corruption detected because bad crc file