[infinispan-issues] [JBoss JIRA] (ISPN-7322) Improve triangle algorithm: ordering by segment

[Bela Ban (JIRA)]

    [ https://issues.jboss.org/browse/ISPN-7322?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13341504#comment-13341504 ] 

Bela Ban commented on ISPN-7322:
--------------------------------

[~rvansa] No, a batch is delivered by *one* thread (same as a single message). In my experiments, full source ordering beats the overhead created by a high number of threads (context switching and lock contention overhead). Also, if you use OOB messages, you won't reap the benefits of improvements such as JGRP-2143.

[~pruivo] Unmarshalling only has a higher cost if you need to allocate memory. Parsing into pre-created (or fixed) memory buffers should be fast.

> Improve triangle algorithm: ordering by segment
> -----------------------------------------------
>
>                 Key: ISPN-7322
>                 URL: https://issues.jboss.org/browse/ISPN-7322
>             Project: Infinispan
>          Issue Type: Enhancement
>          Components: Core
>            Reporter: Pedro Ruivo
>            Assignee: Pedro Ruivo
>
> Current triangle algorithm uses regular message (FIFO ordered) between the primary owner and backup owners of a key. While it ensures that the backup owners receives the stream of updates in the same order, it makes everything slower since it doesn't allow different keys to be handled in parallel.
> "Triangle unordered" solves this problem by sending OOB messages (not ordered) between the primary and backup. To keep the consistency, Infinispan introduces the TriangleOrderManager that orders the updates based on the segment of the key.
> While it is not as perfect as ordering per key, the segments are static; this removes the complexity and avoids handling the cluster topology changes and key adding/removal while improves the performance.



--
This message was sent by Atlassian JIRA
(v7.2.3#72005)