[
https://issues.apache.org/jira/browse/HBASE-4218?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13172872#comment-13172872
]
Phabricator commented on HBASE-4218:
------------------------------------
mbautin has commented on the revision "[jira] [HBASE-4218] Delta encoding for
keys in HFile".
Replying to a part of the comments. Will post a new version when I am done
going through all the pending comments. Running tests, too.
INLINE COMMENTS
src/main/java/org/apache/hadoop/hbase/HColumnDescriptor.java:93 It is
possible to use two different delta encodings on disk and in the block cache.
So e.g. we could use no delta encoding on disk and only delta-encode in cache.
This is the option that we want to use for testing.
In addition to that, there is a boolean option, DELTA_ENCODING_IN_MEMORY,
probably somewhat confusingly named, that Jacek implemented towards the end of
his internship. This option allows to use encoded scanners. I think this might
be OK if we rename this option to make it less confusing and document all three
of these options.
src/main/java/org/apache/hadoop/hbase/KeyValue.java:2020 Done.
src/main/java/org/apache/hadoop/hbase/KeyValue.java:153 Done.
src/main/java/org/apache/hadoop/hbase/KeyValue.java:2036 Done.
src/main/java/org/apache/hadoop/hbase/KeyValue.java:2130 commonPrefix does
include the rowkey portion, but it is OK to pass zero as commonPrefix at line
2051, because this function will not compare the row anyway. I modified the
documentation and got rid of passing lrowlength and rrowlength to this
function, replacing them by only one parameter, because they are always equal.
src/main/java/org/apache/hadoop/hbase/io/deltaencoder/BufferedDeltaEncoder.java:443
Moved the above methods to ByteBufferUtils.
src/main/java/org/apache/hadoop/hbase/io/deltaencoder/BufferedDeltaEncoder.java:470
Nice catch! Fixed this (also made sure that newKeyBufferLength is set to at
least 1).
src/main/java/org/apache/hadoop/hbase/io/deltaencoder/BufferedDeltaEncoder.java:475
Yes, nice catch. Added a unit test.
src/main/java/org/apache/hadoop/hbase/io/deltaencoder/BufferedDeltaEncoder.java:635
Yes, seems like a bug. Fixed.
REVISION DETAIL
https://reviews.facebook.net/D447
> Delta Encoding of KeyValues (aka prefix compression)
> -----------------------------------------------------
>
> Key: HBASE-4218
> URL: https://issues.apache.org/jira/browse/HBASE-4218
> Project: HBase
> Issue Type: Improvement
> Components: io
> Affects Versions: 0.94.0
> Reporter: Jacek Migdal
> Assignee: Mikhail Bautin
> Labels: compression
> Attachments: 0001-Delta-encoding-fixed-encoded-scanners.patch,
> D447.1.patch, D447.2.patch, D447.3.patch, D447.4.patch, D447.5.patch,
> D447.6.patch, D447.7.patch, D447.8.patch,
> Delta_encoding_with_memstore_TS.patch, open-source.diff
>
>
> A compression for keys. Keys are sorted in HFile and they are usually very
> similar. Because of that, it is possible to design better compression than
> general purpose algorithms,
> It is an additional step designed to be used in memory. It aims to save
> memory in cache as well as speeding seeks within HFileBlocks. It should
> improve performance a lot, if key lengths are larger than value lengths. For
> example, it makes a lot of sense to use it when value is a counter.
> Initial tests on real data (key length = ~ 90 bytes , value length = 8 bytes)
> shows that I could achieve decent level of compression:
> key compression ratio: 92%
> total compression ratio: 85%
> LZO on the same data: 85%
> LZO after delta encoding: 91%
> While having much better performance (20-80% faster decompression ratio than
> LZO). Moreover, it should allow far more efficient seeking which should
> improve performance a bit.
> It seems that a simple compression algorithms are good enough. Most of the
> savings are due to prefix compression, int128 encoding, timestamp diffs and
> bitfields to avoid duplication. That way, comparisons of compressed data can
> be much faster than a byte comparator (thanks to prefix compression and
> bitfields).
> In order to implement it in HBase two important changes in design will be
> needed:
> -solidify interface to HFileBlock / HFileReader Scanner to provide seeking
> and iterating; access to uncompressed buffer in HFileBlock will have bad
> performance
> -extend comparators to support comparison assuming that N first bytes are
> equal (or some fields are equal)
> Link to a discussion about something similar:
> http://search-hadoop.com/m/5aqGXJEnaD1/hbase+windows&subj=Re+prefix+compression
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators:
https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa
For more information on JIRA, see: http://www.atlassian.com/software/jira
