[
https://issues.apache.org/jira/browse/HDFS-13056?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Dennis Huo updated HDFS-13056:
------------------------------
Release Note:
<!-- markdown -->
This feature adds a new `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 or local files.
Without this option, HDFS FileChecksums are computed as "MD5 of MD5 of CRC"
across chunks and blocks, so that the checksums are dependent on the configured
chunk and block sizes, rather than only reflecting the logical file contents.
The feature works by changing the way individual chunk and block checksums are
combined at invocation time, so doesn't depend on or affect any persistent
metadata; it can be used in-place in existing HDFS deployments without any
recomputation of block metadata.
This option can be enabled or disabled at the granularity of individual client
calls by setting the new configuration option `dfs.checksum.combine.mode` to
`COMPOSITE_CRC`:
hdfs dfs -Ddfs.checksum.combine.mode=COMPOSITE_CRC -checksum
hdfs:///tmp/foo.txt
By default, this configuration option is set to `MD5MD5CRC`, which preserves
the fully backwards-compatible existing behavior of HDFS.
Note that this "combine mode" is orthogonal to the existing `dfs.checksum.type`
configuration option, which configures the checksum algorithm at the level of
individual chunk crcs that are stored in block metadata, where `CRC32`
specifies the legacy "gzip" CRC polynomial, and `CRC32C` specifies the
"Castagnoli" polynomial. The composite CRC mode will adhere to the checksum
type stored in the block metadata; if an HDFS instance stores a mixture of
files with `CRC32` and `CRC32C` types, the list of FileChecksums with
`COMPOSITE_CRC` will likewise produce the corresponding mixture of
`COMPOSITE-CRC32` and `COMPOSITE-CRC32C` types.
> 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
> Fix For: 3.2.0, 3.1.1
>
> 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-2.8.005.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,
> HDFS-13056.008.patch, HDFS-13056.009.patch, HDFS-13056.010.patch,
> HDFS-13056.011.patch, HDFS-13056.012.patch, HDFS-13056.013.patch,
> HDFS-13056.014.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
(v7.6.3#76005)
---------------------------------------------------------------------
To unsubscribe, e-mail: hdfs-issues-unsubscr...@hadoop.apache.org
For additional commands, e-mail: hdfs-issues-h...@hadoop.apache.org
