You are viewing a plain text version of this content. The canonical link for it is here.
Posted to issues@ignite.apache.org by "Ryabov Dmitrii (JIRA)" <ji...@apache.org> on 2018/07/16 10:56:00 UTC
[jira] [Commented] (IGNITE-4435) Discovery data distribution
process optimization
[ https://issues.apache.org/jira/browse/IGNITE-4435?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16545063#comment-16545063 ]
Ryabov Dmitrii commented on IGNITE-4435:
----------------------------------------
[~sergey-chugunov], looks like ticket wasn't finished - {{filterDuplicatedData}} wasn't removed and still contains [TODO|https://github.com/apache/ignite/blob/0d82f1004c72dd3881e453e555e9803ecd889c74/modules/core/src/main/java/org/apache/ignite/spi/discovery/tcp/internal/DiscoveryDataPacket.java#L324] comment with link to this ticket.
> Discovery data distribution process optimization
> ------------------------------------------------
>
> Key: IGNITE-4435
> URL: https://issues.apache.org/jira/browse/IGNITE-4435
> Project: Ignite
> Issue Type: Improvement
> Components: general
> Reporter: Sergey Chugunov
> Priority: Major
>
> h2. Notes
> Previously upon new node joining topology grid components on each node has been collecting discovery data independently from the same components on other nodes.
> This lead to a situation when *TcpDiscoveryNodeAddedMessage* and *TcpDiscoveryNodeAddFinishedMessage* carried a lot of duplicated data wasting network traffic.
> Refactoring done within [IGNITE-4157|https://issues.apache.org/jira/browse/IGNITE-4157] introduced a way to reduce duplication, but only *GridMarshallerMappingProcessor* component took advantage of it.
> h2. Acceptance Criteria
> * Identify all components needed to be optimized in terms of discovery data collection.
> * Refactor them to reduce duplication of discovery data.
> * Remove *filterDuplicatedData* method as obsolete from *DiscoveryDataContainer* when refactoring is done.
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)