[
https://issues.apache.org/jira/browse/DRILL-4237?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15195956#comment-15195956
]
ASF GitHub Bot commented on DRILL-4237:
---------------------------------------
GitHub user chunhui-shi opened a pull request:
https://github.com/apache/drill/pull/430
DRILL-4237 DRILL-4478 fully implement hash to use murmur3 and corresp…
Since we have to have our own implementations of hash functions, UNIT tests
should guarantee the correctness of implementation. Due to hash function is
critical to distributed SQL engine's performance, make changes to allow future
adding/testing new hash function to be smooth.
1, Implement Murmur3 to work on DrillBuf and long(double), both avoid
object or extra buffer creation.
2, fix xxHash
3, Point hash functions in hash32, hash64, hash32AsDouble and
hash64AsDouble to HashHelper.hash32|64
4, Test Correctness of the hash functions by comparing with original c
implementation's generated hash results.
Testing vectors are generated from https://github.com/chunhui-shi/smhasher
by running these commands after build:
gethash Murmur3F --json=MurmurHash3
gethash xxHash64 --json=XXHash
5, Adding tests to cover Hash**WithSeed.
6, Fix DRILL-4478 which prevent binar_string function working correctly.
You can merge this pull request into a Git repository by running:
$ git pull https://github.com/chunhui-shi/drill hash
Alternatively you can review and apply these changes as the patch at:
https://github.com/apache/drill/pull/430.patch
To close this pull request, make a commit to your master/trunk branch
with (at least) the following in the commit message:
This closes #430
----
commit 2210279618dee857df32f28c647ee8ec8409a44a
Author: chunhui-shi <[email protected]>
Date: 2016-03-15T17:36:54Z
DRILL-4237 DRILL-4478 fully implement hash to use murmur3 and correspondent
unit tests
----
> Skew in hash distribution
> -------------------------
>
> Key: DRILL-4237
> URL: https://issues.apache.org/jira/browse/DRILL-4237
> Project: Apache Drill
> Issue Type: Bug
> Components: Functions - Drill
> Affects Versions: 1.4.0
> Reporter: Aman Sinha
> Assignee: Chunhui Shi
>
> Apparently, the fix in DRILL-4119 did not fully resolve the data skew issue.
> It worked fine on the smaller sample of the data set but on another sample of
> the same data set, it still produces skewed values - see below the hash
> values which are all odd numbers.
> {noformat}
> 0: jdbc:drill:zk=local> select columns[0], hash32(columns[0]) from `test.csv`
> limit 10;
> +-----------------------------------+--------------+
> | EXPR$0 | EXPR$1 |
> +-----------------------------------+--------------+
> | f71aaddec3316ae18d43cb1467e88a41 | 1506011089 |
> | 3f3a13bb45618542b5ac9d9536704d3a | 1105719049 |
> | 6935afd0c693c67bba482cedb7a2919b | -18137557 |
> | ca2a938d6d7e57bda40501578f98c2a8 | -1372666789 |
> | fab7f08402c8836563b0a5c94dbf0aec | -1930778239 |
> | 9eb4620dcb68a84d17209da279236431 | -970026001 |
> | 16eed4a4e801b98550b4ff504242961e | 356133757 |
> | a46f7935fea578ce61d8dd45bfbc2b3d | -94010449 |
> | 7fdf5344536080c15deb2b5a2975a2b7 | -141361507 |
> | b82560a06e2e51b461c9fe134a8211bd | -375376717 |
> +-----------------------------------+--------------+
> {noformat}
> This indicates an underlying issue with the XXHash64 java implementation,
> which is Drill's implementation of the C version. One of the key difference
> as pointed out by [~jnadeau] was the use of unsigned int64 in the C version
> compared to the Java version which uses (signed) long. I created an XXHash
> version using com.google.common.primitives.UnsignedLong. However,
> UnsignedLong does not have bit-wise operations that are needed for XXHash
> such as rotateLeft(), XOR etc. One could write wrappers for these but at
> this point, the question is: should we think of an alternative hash function
> ?
> The alternative approach could be the murmur hash for numeric data types that
> we were using earlier and the Mahout version of hash function for string
> types
> (https://github.com/apache/drill/blob/master/exec/java-exec/src/main/java/org/apache/drill/exec/expr/fn/impl/HashHelper.java#L28).
> As a test, I reverted to this function and was getting good hash
> distribution for the test data.
> I could not find any performance comparisons of our perf tests (TPC-H or DS)
> with the original and newer (XXHash) hash functions. If performance is
> comparable, should we revert to the original function ?
> As an aside, I would like to remove the hash64 versions of the functions
> since these are not used anywhere.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
