ignite-issues mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Dmitry Kiriy (JIRA)" <j...@apache.org>
Subject [jira] [Updated] (IGNITE-10703) Queue drainTo should be batched operation
Date Sat, 15 Dec 2018 08:05:00 GMT

     [ https://issues.apache.org/jira/browse/IGNITE-10703?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]

Dmitry Kiriy updated IGNITE-10703:
----------------------------------
    Description: 
Currently drainTo(Collection<T> to, int amount) implemented as cyclic call of poll()
method, which is ineffective.

Need to be implemented as truly batch operation - trying to read required amount of data at
once.

current design prevent high-performance data transfer cross queue.

P.S. addAll implemented as batch, so together with right drainTo implementation, it will be
complete high-performance batched read&write queue solution

  was:
Currently drainTo(Collection<T> to, int amount) implemented as cyclic call of poll()
method, which is ineffective.

Need to be implemented as truly batch operation - trying to read required amount of data at
once.

current design prevent high-performance data transfer cross queue.

P.S. addAll implemented as batch, so together with right drainTo implementation, it will be
complete high-performance batched read&write solution


> Queue drainTo should be batched operation
> -----------------------------------------
>
>                 Key: IGNITE-10703
>                 URL: https://issues.apache.org/jira/browse/IGNITE-10703
>             Project: Ignite
>          Issue Type: Improvement
>          Components: data structures
>    Affects Versions: 2.7
>            Reporter: Dmitry Kiriy
>            Priority: Major
>              Labels: performance
>   Original Estimate: 168h
>  Remaining Estimate: 168h
>
> Currently drainTo(Collection<T> to, int amount) implemented as cyclic call of poll()
method, which is ineffective.
> Need to be implemented as truly batch operation - trying to read required amount of data
at once.
> current design prevent high-performance data transfer cross queue.
> P.S. addAll implemented as batch, so together with right drainTo implementation, it will
be complete high-performance batched read&write queue solution



--
This message was sent by Atlassian JIRA
(v7.6.3#76005)

Mime
View raw message