[ 
https://issues.apache.org/jira/browse/IGNITE-18637?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
 ]

Aleksey Plekhanov updated IGNITE-18637:
---------------------------------------
    Labels: important ise  (was: ise)

> Java thin client: Connections balancing
> ---------------------------------------
>
>                 Key: IGNITE-18637
>                 URL: https://issues.apache.org/jira/browse/IGNITE-18637
>             Project: Ignite
>          Issue Type: Improvement
>          Components: thin client
>            Reporter: Aleksey Plekhanov
>            Assignee: Aleksey Plekhanov
>            Priority: Major
>              Labels: important, ise
>             Fix For: 2.15
>
>          Time Spent: 20m
>  Remaining Estimate: 0h
>
> Since Ignite 2.12 java thin client always connects to the first address from 
> the provided address list. In typical configuration this leads to overloading 
> an one server node and underloading other server nodes. 
> To improve workload distribution we can:
>  # Choose random first channel from the range of addresses with the same 
> (minimal) port.
>  # If partition awareness is enabled use random open channel instead of 
> default channel for operations which can't be mapped to a certain node. 



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

Reply via email to