You are viewing a plain text version of this content. The canonical link for it is here.
Posted to issues@ignite.apache.org by "Andrey N. Gura (Jira)" <ji...@apache.org> on 2023/05/15 11:44:00 UTC

[jira] [Updated] (IGNITE-16677) [Compute Grid] Implement broadcast method of IgniteCompute interface

     [ https://issues.apache.org/jira/browse/IGNITE-16677?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Andrey N. Gura updated IGNITE-16677:
------------------------------------
    Labels: iep-100 ignite-3  (was: ignite-3)

> [Compute Grid] Implement broadcast method of IgniteCompute interface
> --------------------------------------------------------------------
>
>                 Key: IGNITE-16677
>                 URL: https://issues.apache.org/jira/browse/IGNITE-16677
>             Project: Ignite
>          Issue Type: Task
>          Components: compute
>            Reporter: Sergey Chugunov
>            Assignee: Roman Puchkovskiy
>            Priority: Major
>              Labels: iep-100, ignite-3
>             Fix For: 3.0.0-alpha5
>
>          Time Spent: 50m
>  Remaining Estimate: 0h
>
> The following method should be implemented:
> {code:java}
> <R> Map<ClusterNode, CompletableFuture<R>> broadcast(Set<ClusterNode> nodes, Class<? extends ComputeJob<R>> cls, Object … args);
> {code}
> On the first phase implementation doesn't need to support failover.
> Each future from returned map represents the result of execution on the corresponding node and is independent from other futures.
> It is a deliberate choice to make this API simpler but harder to work with on the first phase.
> More discussions are needed to make it more convenient for the user and reduce amount of boilerplate code needed.



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