[ 
https://issues.apache.org/jira/browse/HIVE-19820?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
 ]

Sergey Shelukhin updated HIVE-19820:
------------------------------------
       Resolution: Fixed
    Fix Version/s: txnstats
           Status: Resolved  (was: Patch Available)

Committed to branch. I've fixed all the failed tests that were not already 
failing on the branch, from the last run.
If I've broken more tests in process, they will be fixed together with the rest 
of the broken tests on the branch.

[~ekoifman] there's one problem that's kind of papered over (I made an 
exception in the code)...
Currently we generate write ID when converting to MM tables. However, if I add 
the same path (generate write ID in Driver acquireLocks using the normal 
process) when converting to full ACID tables, some tests fail because it 
generates a normal write ID instead of the magical write ID that is currently 
used (10000001 or smth like that) for conversion.
Why cannot full ACID conversion use normal write IDs? If not, where is this 
magic write ID generated, and can we instead create it via normal means in 
acquireLocks?

> add ACID stats support to background stats updater and fix bunch of edge 
> cases found in SU tests
> ------------------------------------------------------------------------------------------------
>
>                 Key: HIVE-19820
>                 URL: https://issues.apache.org/jira/browse/HIVE-19820
>             Project: Hive
>          Issue Type: Sub-task
>            Reporter: Sergey Shelukhin
>            Assignee: Sergey Shelukhin
>            Priority: Major
>             Fix For: txnstats
>
>         Attachments: HIVE-19820.01-master-txnstats.patch, 
> HIVE-19820.02-master-txnstats.patch, HIVE-19820.03-master-txnstats.patch, 
> HIVE-19820.04-master-txnstats.patch, HIVE-19820.05.patch, 
> branch-19820.02.nogen.patch, branch-19820.03.nogen.patch, 
> branch-19820.04.nogen.patch, branch-19820.nogen.patch
>
>
> Follow-up from HIVE-19418.
> Right now it checks whether stats are valid in an old-fashioned way... and 
> also gets ACID state, and discards it without using.
> When ACID stats are implemented, ACID state needs to be used to do 
> version-aware valid stats checks.



--
This message was sent by Atlassian JIRA
(v7.6.3#76005)

Reply via email to