[ 
https://issues.apache.org/jira/browse/HBASE-18946?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16278985#comment-16278985
 ] 

Anoop Sam John commented on HBASE-18946:
----------------------------------------

bq.Can we give the replicas as a group to the LB and have it queue them as a 
group? I
Yes same thing I ask/suggest from long time.. As discussed with Ram, we have 
some issue in there with the usage of Procedure way.  But my call there would 
be , if this need a change in the Procedure f/w on the sub procedure ways or 
so, lets do that.
Creating the assign individual procedures for each of the replicas at 1st step 
and then ask LB is the issue maker.     In assign procedure, first we add entry 
to META.  What if there we dont have the replica regions(objects) been created 
by that time?  I think it may be ok as we dont have row per replica region in 
META. It is just one row only for a region and as per the replicas some new 
columns (cf:Q) comes in only.   Then while contacting LB, we can create the 
replica regions and pass it as a List to get the servers for regions.  Already 
the LB APIs allow to pass a list of regions so as to assign servers for each.   
Now the Q is already each of the region assign is a sub procedure within the 
Table create procedure.   Can we again make sub procedures for each of the 
replica regions under a region sub procedure?  [~stack]..   Not knowing the 
internals of Procedure f/w that well ...  But will this direction help?

> Stochastic load balancer assigns replica regions to the same RS
> ---------------------------------------------------------------
>
>                 Key: HBASE-18946
>                 URL: https://issues.apache.org/jira/browse/HBASE-18946
>             Project: HBase
>          Issue Type: Bug
>    Affects Versions: 2.0.0-alpha-3
>            Reporter: ramkrishna.s.vasudevan
>            Assignee: ramkrishna.s.vasudevan
>             Fix For: 2.0.0-beta-1
>
>         Attachments: HBASE-18946.patch, HBASE-18946.patch, 
> HBASE-18946_2.patch, HBASE-18946_2.patch, HBASE-18946_simple_7.patch, 
> HBASE-18946_simple_8.patch, TestRegionReplicasWithRestartScenarios.java
>
>
> Trying out region replica and its assignment I can see that some times the 
> default LB Stocahstic load balancer assigns replica regions to the same RS. 
> This happens when we have 3 RS checked in and we have a table with 3 
> replicas. When a RS goes down then the replicas being assigned to same RS is 
> acceptable but the case when we have enough RS to assign this behaviour is 
> undesirable and does not solve the purpose of replicas. 
> [~huaxiang] and [~enis]. 



--
This message was sent by Atlassian JIRA
(v6.4.14#64029)

Reply via email to