[
https://issues.apache.org/jira/browse/HIVE-18814?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16422816#comment-16422816
]
Alan Gates commented on HIVE-18814:
-----------------------------------
There are musings in several possible solutions to this problem in the
description but no indication which was chosen. Could you describe the
approach you took to help me better understand the patch?
> Support Add Partition For Acid tables
> -------------------------------------
>
> Key: HIVE-18814
> URL: https://issues.apache.org/jira/browse/HIVE-18814
> Project: Hive
> Issue Type: New Feature
> Components: Transactions
> Reporter: Eugene Koifman
> Assignee: Eugene Koifman
> Priority: Major
> Attachments: HIVE-18814.01.patch, HIVE-18814.02.patch,
> HIVE-18814.03.patch, HIVE-18814.04.patch
>
>
> [https://cwiki.apache.org/confluence/display/Hive/LanguageManual%2BDDL#LanguageManualDDL-AddPartitions]
> Add Partition command creates a {{Partition}} metadata object and sets the
> location to the directory containing data files.
> In current master (Hive 3.0), Add partition on an acid table doesn't fail and
> at read time the data is decorated with row__id but the original transaction
> is 0. I suspect in earlier Hive versions this will throw or return no data.
> Since this new partition didn't have data before, assigning txnid:0 isn't
> going to generate duplicate IDs but it could violate Snapshot Isolation in
> multi stmt txns. Suppose txnid:7 runs {{select * from T}}. Then txnid:8
> adds a partition to T. Now if txnid:7 runs the same query again, it will see
> the data in the new partition.
> This can't be release like this since a delete on this data (added via Add
> partition) will use row_ids with txnid:0 so a later upgrade that sees
> un-compacted may generate row_ids with different txnid (assuming this is
> fixed by then)
>
> One option is follow Load Data approach and create a new delta_x_x/ and
> move/copy the data there.
>
> Another is to allocate a new writeid and save it in Partition metadata. This
> could then be used to decorate data with ROW__IDs. This avoids move/copy but
> retains data "outside" of the table tree which make it more likely that this
> data will be modified in some way which can really break things if done after
> and SQL update/delete on this data have happened.
>
> It performs no validations on add (except for partition spec) so any file
> with any format can be added. It allows add to bucketed tables as well.
> Seems like a very dangerous command. Maybe a better option is to block it
> and advise using Load Data. Alternatively, make this do Add partition
> metadata op followed by Load Data.
>
>
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
