[MDEV-21291] Support Parallel Query Execution Created: 2019-12-11  Updated: 2020-11-10  Resolved: 2020-11-10

Status: Closed
Project: MariaDB Server
Component/s: Optimizer
Fix Version/s: N/A

Type: Task Priority: Major
Reporter: Matt Wegrzyn Assignee: Unassigned
Resolution: Duplicate Votes: 0
Labels: None

Issue Links:
Duplicate
is duplicated by MDEV-6096 Ideas about parallel query execution Open
Relates
relates to MCOL-2262 Design efficient methods for interact... Closed
relates to MDEV-18368 MySQL already can do parallel queries... Closed
relates to MDEV-18705 Parallel index range scan Open

 Description   

Support Parallel Query Execution

Come on, a decade over due.

Entire column store is basically pointless. Just support parallel query execution.



 Comments   
Comment by Andrew Hutchings (Inactive) [ 2019-12-12 ]

Hi Matt,

Thanks for creating this ticket.

I think first of all it needs to be defined what your view of parallel query execution really is. What parts would you like parallelised and what do you see the advantages are of doing this?

Typical queries run relatively quickly and you can run queries in parallel in most MariaDB engines. Those queries that don't fit this criteria usually run relatively quickly in ColumnStore.

You have shared an opinion that ColumnStore is pointless but haven't stated what you don't like about it, without this information we cannot improve on it. We are working towards tighter integration between MariaDB and ColumnStore which you will start to see over the coming months. You are welcome to open feature requests on this (I recommend using the MCOL project for that in Jira).

If there are use cases where parellisation hits a sweet spot that cannot be hit be ColumnStore, please describe your use case in much more detail and the MariaDB & ColumnStore engineering teams can weigh up the best ways of supporting this.


Andrew Hutchings
Software Engineering Manager for ColumnStore.

Generated at Thu Feb 08 09:06:02 UTC 2024 using Jira 8.20.16#820016-sha1:9d11dbea5f4be3d4cc21f03a88dd11d8c8687422.