[
https://issues.apache.org/jira/browse/DRILL-6835?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16698042#comment-16698042
]
Pritesh Maker commented on DRILL-6835:
--------------------------------------
Hi [~arina]
I just got around to reading the proposal. I thought it was a well-researched
and written proposal - nice work! It's an excellent way to address the common
issues that users have with schema change exceptions.
There are several big items in this proposal and I see some discussions on how
to phase in the efforts. Do you have some thoughts on this? Maybe we can
discuss it in a hangout as well. My initial thoughts are as follows: * [Phase
1] Schema provision using File/ Table function (could be broken down further
into the File approach first followed by a table function. Eventually,
integrate with
* [Phase 2] Taking advantage of the schema during execution (for example, push
casts into scans). I am sure there must be other opportunities for the
optimizer/ runtime to work differently when the schema is known.
* [Phase 3] Analyze the data for schema disocvery and recommended schema file.
Thanks,
Pritesh
> Schema Provision using File / Table Function
> --------------------------------------------
>
> Key: DRILL-6835
> URL: https://issues.apache.org/jira/browse/DRILL-6835
> Project: Apache Drill
> Issue Type: New Feature
> Reporter: Arina Ielchiieva
> Assignee: Arina Ielchiieva
> Priority: Major
> Fix For: 1.16.0
>
>
> Schema Provision using File / Table Function design document:
> https://docs.google.com/document/d/1mp4egSbNs8jFYRbPVbm_l0Y5GjH3HnoqCmOpMTR_g4w/edit?usp=sharing
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
