[ 
https://issues.apache.org/jira/browse/ARTEMIS-2716?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17084722#comment-17084722
 ] 

Francesco Nigro commented on ARTEMIS-2716:
------------------------------------------

The previous objectives could be translated, with more details, in:
# preserve compatibility at configuration level (ok to add new things, trying 
hard to not deprecate anything)
# it should cover the same cases of the original quorum vote
# it should cover some additional case ie 1 pair (+ 1 witness "node" - not a 
full broker)

At a lower level:
# clients shouldn't change: it means that at a lower level it will reuse the 
same "gossip protocol" based on Topology propagation used now
# the mechanism used to perform the backup pairing with live (that include both 
the discovery process and the synchronization with it) shouldn't change

A proposal of the next steps to be followed to get this is:
# abstract away the current quorum vote: it requires extra-care because the
logic is mixed together with the replication/clustering behaviour
# refactor it in order to separate election phase and cluster member states
# design a new generic API that should just cover the higher-level 
functionalities provided by the previous implementation
# choose/discuss a possible (consensus algorithm) provider that could be used 
to implement a first POC (maybe as first step to get a RI)
# implement a RI version

> Implements pluggable Quorum Vote
> --------------------------------
>
>                 Key: ARTEMIS-2716
>                 URL: https://issues.apache.org/jira/browse/ARTEMIS-2716
>             Project: ActiveMQ Artemis
>          Issue Type: New Feature
>            Reporter: Francesco Nigro
>            Priority: Major
>
> This task aim to ideliver a new Quorum Vote mechanism for artemis with the 
> objectives:
> # to make it pluggable
> # to cleanly separate the election phase and the cluster member states
> # to simplify most common setups in both amount of configuration and 
> requirements (eg "witness" nodes could be implemented to support single 
> master-slave pairs)
> Post-actions to help people adopt it, but need to be thought upfront:
> # a clean upgrade path for current HA replication users
> # deprecate or integrate the current HA replication into the new version



--
This message was sent by Atlassian Jira
(v8.3.4#803005)

Reply via email to