[ 
https://issues.apache.org/jira/browse/IGNITE-17412?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
 ]

Yury Gerzhedovich updated IGNITE-17412:
---------------------------------------
    Description: 
Exchange of sql fragments is havy and complex operation due to include 
serialize/deserialize operations and recreate execution context for fragment

For prepared statements we could avoid extra sending of wury fragments and have 
separate cache SQL fragments on map nodes. As for example it could be some id 
which will be generate during first send/execute of fragment and using for all 
next execution for the same prepared statements. Let's try inroduce the 
optimization and do some benchmarks to show performance increase.

  was:
Exchange of sql fragments is havy and complex operation due to include 
serialize/deserialize operations and recreate execution context for fragment

For prepared statements we could avoid extra sending of wury fragments and have 
separate cache SQL fragments on map nodes. As for example it could be some id 
which will be generate during first send/execute of fragment and using for all 
next execution for the same prepared statements. Let's try inrofuce the 
optimization and do some benchmarks to show performance increase.


> Optimize exchange sql fragments for prepared statements
> -------------------------------------------------------
>
>                 Key: IGNITE-17412
>                 URL: https://issues.apache.org/jira/browse/IGNITE-17412
>             Project: Ignite
>          Issue Type: Improvement
>            Reporter: Yury Gerzhedovich
>            Priority: Major
>              Labels: ignite-3
>
> Exchange of sql fragments is havy and complex operation due to include 
> serialize/deserialize operations and recreate execution context for fragment
> For prepared statements we could avoid extra sending of wury fragments and 
> have separate cache SQL fragments on map nodes. As for example it could be 
> some id which will be generate during first send/execute of fragment and 
> using for all next execution for the same prepared statements. Let's try 
> inroduce the optimization and do some benchmarks to show performance increase.



--
This message was sent by Atlassian Jira
(v8.20.10#820010)

Reply via email to