You are viewing a plain text version of this content. The canonical link for it is here.

Posted to commits@cassandra.apache.org by "Vijay (JIRA)" <ji...@apache.org> on 2011/03/25 01:48:06 UTC

[jira] [Updated] (CASSANDRA-2300) Enhance Node Bootstrapping, while adding more than 1 nodes to the cluster

     [ https://issues.apache.org/jira/browse/CASSANDRA-2300?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Vijay updated CASSANDRA-2300:
-----------------------------

    Attachment: 001-2300-BootStrapper.patch
                002-2300-BootstrapperVerb.patch
                003-2300-BootStrapperTest.patch

This patch makes the assignment of the token in a range Atomic... It would be nice to have the same order of the seed nodes in the yaml file else.. though it wont break anything.

> Enhance Node Bootstrapping, while adding more than 1 nodes to the cluster
> -------------------------------------------------------------------------
>
>                 Key: CASSANDRA-2300
>                 URL: https://issues.apache.org/jira/browse/CASSANDRA-2300
>             Project: Cassandra
>          Issue Type: New Feature
>          Components: Core
>    Affects Versions: 0.8
>         Environment: JVM
>            Reporter: Vijay
>            Assignee: Vijay
>            Priority: Minor
>         Attachments: 001-2300-BootStrapper.patch, 002-2300-BootstrapperVerb.patch, 003-2300-BootStrapperTest.patch
>
>
> Problem: When we add more than one node to the cluster, nodes tends to get the same token range's.... currently we have to wait till the node starts to stream the data (Stage 3 as listed below)
> There are 3 stages while the node trys to bootstrap
> 1) It understands the Ring
> 2) It finds a token in the ring and announce to the cluster
> 3) Start to stream the data.
> Possible Solution: If we can move the stage 2 from the bootstrapping node to the seed node's, when it starts the boot strapping it will be easier for the other nodes who are joining. 
> This will reduce the time which is needed for us to add nodes to the clusters. This will also reduce the manual work of assigning the tokens when doubling the cluster and opens up possibilities of future enhancements to the token assignment.

--
This message is automatically generated by JIRA.
For more information on JIRA, see: http://www.atlassian.com/software/jira