[ 
https://issues.apache.org/jira/browse/HBASE-6721?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15162687#comment-15162687
 ] 

Francis Liu commented on HBASE-6721:
------------------------------------

{quote}
Yeah hadoop can do whatever they want. I would prefer that HBase is consistent 
whenever interacting with jmx so when we make changes in the future it's all 
moving from the same base.
{quote}
Took a look at this further it may be possible to reuse tags to publish 
metadata in some form but I would need to add APIs to the base (BaseSource?). 
Tho IMHO we could probably do better by providing a common base on which to 
publish metadata via jmx instead of retrofitting it into hadoop-metrics which 
looking at the direction hadoop went is not how they intend metrics subsystem 
to be used. So instead of addressing this in this patch I propose we create two 
new jiras: 1. create apis to publish metadata via jmx, 2. create jira to 
leverage #1 and publish region server groups metadata. Thoughts?

{quote}
Both formats have been able to be parsed for a long time. So if the test needs 
to be changed to pass then that's breaking change. If not then we shouldn't 
have changes that aren't needed in a patch this large.
{quote}
I'm not sure how the format used by TestMasterStatusServlet: 
rootserver:123,12345 is actually a valid format. It seems like a typo/bug to 
me. For a few reasons:

1. The format is not documented in the javadoc
2. If you look at how the format mentioned is parsed the components would come 
out as host=rootserver:123/port=12345/startcode=12345. Which doesn't make much 
sense? Also ':' is not a valid character for a hostname according to RFC. 

What uses this format? Let me know what I'm missing. On a related note if we 
support this format we really can't use HostAndPort as it assumes ':' will 
delimit a port and will throw an exception if it finds it in the hostname.

On a related note another behavior change since the introduction of HostAndPort 
is that '-1' can no longer be used as a port number (or any negative number). 
Would this be a breaking change? Let me know how you'd like to go about this 
change. 



> RegionServer Group based Assignment
> -----------------------------------
>
>                 Key: HBASE-6721
>                 URL: https://issues.apache.org/jira/browse/HBASE-6721
>             Project: HBase
>          Issue Type: New Feature
>            Reporter: Francis Liu
>            Assignee: Francis Liu
>              Labels: hbase-6721
>         Attachments: 6721-master-webUI.patch, HBASE-6721 
> GroupBasedLoadBalancer Sequence Diagram.xml, HBASE-6721-DesigDoc.pdf, 
> HBASE-6721-DesigDoc.pdf, HBASE-6721-DesigDoc.pdf, HBASE-6721-DesigDoc.pdf, 
> HBASE-6721_0.98_2.patch, HBASE-6721_10.patch, HBASE-6721_11.patch, 
> HBASE-6721_12.patch, HBASE-6721_13.patch, HBASE-6721_14.patch, 
> HBASE-6721_15.patch, HBASE-6721_8.patch, HBASE-6721_9.patch, 
> HBASE-6721_9.patch, HBASE-6721_94.patch, HBASE-6721_94.patch, 
> HBASE-6721_94_2.patch, HBASE-6721_94_3.patch, HBASE-6721_94_3.patch, 
> HBASE-6721_94_4.patch, HBASE-6721_94_5.patch, HBASE-6721_94_6.patch, 
> HBASE-6721_94_7.patch, HBASE-6721_98_1.patch, HBASE-6721_98_2.patch, 
> HBASE-6721_hbase-6721_addendum.patch, HBASE-6721_trunk.patch, 
> HBASE-6721_trunk.patch, HBASE-6721_trunk.patch, HBASE-6721_trunk1.patch, 
> HBASE-6721_trunk2.patch, balanceCluster Sequence Diagram.svg, 
> hbase-6721-v15-branch-1.1.patch, hbase-6721-v16.patch, hbase-6721-v17.patch, 
> hbase-6721-v18.patch, hbase-6721-v19.patch, hbase-6721-v20.patch, 
> hbase-6721-v21.patch, hbase-6721-v22.patch, hbase-6721-v23.patch, 
> hbase-6721-v25.patch, immediateAssignments Sequence Diagram.svg, 
> randomAssignment Sequence Diagram.svg, retainAssignment Sequence Diagram.svg, 
> roundRobinAssignment Sequence Diagram.svg
>
>
> In multi-tenant deployments of HBase, it is likely that a RegionServer will 
> be serving out regions from a number of different tables owned by various 
> client applications. Being able to group a subset of running RegionServers 
> and assign specific tables to it, provides a client application a level of 
> isolation and resource allocation.
> The proposal essentially is to have an AssignmentManager which is aware of 
> RegionServer groups and assigns tables to region servers based on groupings. 
> Load balancing will occur on a per group basis as well. 
> This is essentially a simplification of the approach taken in HBASE-4120. See 
> attached document.



--
This message was sent by Atlassian JIRA
(v6.3.4#6332)

Reply via email to