[
https://issues.apache.org/jira/browse/HIVE-15335?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15751776#comment-15751776
]
Owen O'Malley commented on HIVE-15335:
--------------------------------------
Speeding up the decimal vectorization is insanely important. I'm really glad
that you are working on the patch.
{quote}
I have great difficulty accepting that DecimalColumnVector is now a public API.
{quote}
It isn't hard to see:
* vectorized UDF will break
* ORC users that don't use row by row shims will break
Both are documented and public APIs.
I'm -1 to incompatible changes to the storage-api API. You need to add new
methods/classes and deprecate the old ones and remove them at some point down
the road. On the plus side, if you don't make incompatible changes to the lower
levels, you don't need to make a single huge patch and can divide the jira into
more managable patches.
The current DecimalColumnVector is so incredibly broken, that completely
replacing the implementation is great. I haven't looked at the representation,
yet. How did you encode the decimals in your FastDecimalColumnVector? I really
hope it is an array of longs like Teddy was using.
> Fast Decimal
> ------------
>
> Key: HIVE-15335
> URL: https://issues.apache.org/jira/browse/HIVE-15335
> Project: Hive
> Issue Type: Bug
> Components: Hive
> Reporter: Matt McCline
> Assignee: Matt McCline
> Priority: Critical
> Attachments: HIVE-15335.01.patch, HIVE-15335.02.patch,
> HIVE-15335.03.patch, HIVE-15335.04.patch, HIVE-15335.05.patch,
> HIVE-15335.06.patch, HIVE-15335.07.patch, HIVE-15335.08.patch,
> HIVE-15335.09.patch, HIVE-15335.091.patch, HIVE-15335.092.patch,
> HIVE-15335.093.patch
>
>
> Replace HiveDecimal implementation that currently represents the decimal
> internally as a BigDecimal with a faster version that does not allocate extra
> objects
> Replace HiveDecimalWritable implementation with a faster version that has new
> mutable* calls (e.g. mutableAdd, mutableEnforcePrecisionScale, etc) and
> stores the result as a fast decimal instead of a slow byte array containing a
> serialized BigInteger.
> Provide faster ways to serialize/deserialize decimals.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
