Dennis Huo commented on HDFS-13056:

Good catch, thanks for the report, [~ajayydv]. I actually do remember going 
over that shared precondition check and intentionally not supporting it at the 
time since bytesPerCRC propagates out and gets used in preserving file 
attributes (there's a section on file-attribute propagation in the design doc), 
but now on second thought, since this constraint is only being imposed in the 
client-side (doesn't have the same issue in BlockChecksumHelper), it seems 
simple enough to also support mixed-bpc files (though, it's still intended to 
*not* support mixed-checksum-type files where some blocks are CRC32 and others 

I'll still spit out a warning though since the best we can do in this case is 
arbitrarily choose which bpc to default to when attribute preservation includes 
bpc. This will also have the interesting side-effect that whereas distcp 
currently will fail to copy source files with mixed BPC if 
FileAttribute.CHECKSUMTYPE is included as an attribute to preserve, now with 
COMPOSITE_CRC those distcp jobs will succeed, although the mixed-bpc 
characteristics themselves will not actually be preserved.

> Expose file-level composite CRCs in HDFS which are comparable across 
> different instances/layouts
> ------------------------------------------------------------------------------------------------
>                 Key: HDFS-13056
>                 URL: https://issues.apache.org/jira/browse/HDFS-13056
>             Project: Hadoop HDFS
>          Issue Type: New Feature
>          Components: datanode, distcp, erasure-coding, federation, hdfs
>    Affects Versions: 3.0.0
>            Reporter: Dennis Huo
>            Assignee: Dennis Huo
>            Priority: Major
>         Attachments: HDFS-13056-branch-2.8.001.patch, 
> HDFS-13056-branch-2.8.002.patch, HDFS-13056-branch-2.8.003.patch, 
> HDFS-13056-branch-2.8.004.patch, HDFS-13056-branch-2.8.poc1.patch, 
> HDFS-13056.001.patch, HDFS-13056.002.patch, HDFS-13056.003.patch, 
> HDFS-13056.003.patch, HDFS-13056.004.patch, HDFS-13056.005.patch, 
> HDFS-13056.006.patch, HDFS-13056.007.patch, 
> Reference_only_zhen_PPOC_hadoop2.6.X.diff, hdfs-file-composite-crc32-v1.pdf, 
> hdfs-file-composite-crc32-v2.pdf, hdfs-file-composite-crc32-v3.pdf
> FileChecksum was first introduced in 
> [https://issues-test.apache.org/jira/browse/HADOOP-3981] and ever since then 
> has remained defined as MD5-of-MD5-of-CRC, where per-512-byte chunk CRCs are 
> already stored as part of datanode metadata, and the MD5 approach is used to 
> compute an aggregate value in a distributed manner, with individual datanodes 
> computing the MD5-of-CRCs per-block in parallel, and the HDFS client 
> computing the second-level MD5.
> A shortcoming of this approach which is often brought up is the fact that 
> this FileChecksum is sensitive to the internal block-size and chunk-size 
> configuration, and thus different HDFS files with different block/chunk 
> settings cannot be compared. More commonly, one might have different HDFS 
> clusters which use different block sizes, in which case any data migration 
> won't be able to use the FileChecksum for distcp's rsync functionality or for 
> verifying end-to-end data integrity (on top of low-level data integrity 
> checks applied at data transfer time).
> This was also revisited in https://issues.apache.org/jira/browse/HDFS-8430 
> during the addition of checksum support for striped erasure-coded files; 
> while there was some discussion of using CRC composability, it still 
> ultimately settled on hierarchical MD5 approach, which also adds the problem 
> that checksums of basic replicated files are not comparable to striped files.
> This feature proposes to add a "COMPOSITE-CRC" FileChecksum type which uses 
> CRC composition to remain completely chunk/block agnostic, and allows 
> comparison between striped vs replicated files, between different HDFS 
> instances, and possible even between HDFS and other external storage systems. 
> This feature can also be added in-place to be compatible with existing block 
> metadata, and doesn't need to change the normal path of chunk verification, 
> so is minimally invasive. This also means even large preexisting HDFS 
> deployments could adopt this feature to retroactively sync data. A detailed 
> design document can be found here: 
> https://storage.googleapis.com/dennishuo/hdfs-file-composite-crc32-v1.pdf

This message was sent by Atlassian JIRA

To unsubscribe, e-mail: hdfs-issues-unsubscr...@hadoop.apache.org
For additional commands, e-mail: hdfs-issues-h...@hadoop.apache.org

Reply via email to