[
https://issues.apache.org/jira/browse/IGNITE-4106?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15830146#comment-15830146
]
Andrew Mashenkov commented on IGNITE-4106:
------------------------------------------
I create a new PR review due to we resolve some sql issues that was fixed since
I start to resolve this ticket, but Upsource consider that merging with master
branch is a part of PR.
[~sergi.vladykin], I've fixed almost all places you point. Olny one thing I
cant understand is "As discussed earlier: having that we can send multiple
local requests, we have to handle them accordingly. I suspect here we have a
bug."
Do you mean we should send message to local node via ctx.io().send() as we do
for remotes? For local nodes we call message handler directly and the handler
bother about running parallel MapQueries as we do for remote messages.
> SQL: parallelize sql queries over cache local partitions
> --------------------------------------------------------
>
> Key: IGNITE-4106
> URL: https://issues.apache.org/jira/browse/IGNITE-4106
> Project: Ignite
> Issue Type: Improvement
> Components: SQL
> Affects Versions: 1.7
> Reporter: Andrew Mashenkov
> Assignee: Andrew Mashenkov
> Labels: performance
> Fix For: 1.9
>
> Attachments: 1node-4thread.jfr, 4node-1thread.jfr
>
>
> If we run SQL query on cache partitioned over several cluster nodes, it will
> be split into several queries running in parallel. But really we will have
> one thread per query on each node.
> So, for now, to improve SQL query performance we need to run more Ignite
> instances or split caches manually.
> It seems to be better to split local SQL queries over cache partitions, so we
> would be able to parallelize SQL query on every single node and utilize CPU
> more efficiently.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
