[
https://issues.apache.org/jira/browse/HBASE-4755?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13589278#comment-13589278
]
Devaraj Das commented on HBASE-4755:
------------------------------------
Thanks [~jmhsieh] for spending time on the issue. I think I should write a spec
up and we can continue the discussion on that spec as the base. Answers to some
questions/comments:
bq. I'm assuming that HBase provides the list of DN's, likely selected at
region creation time?
Yes (in particular, I am currently only considering pre-split tables, and the
createTable call in the master allocates the regions to particular datanodes).
The regionservers would use this information (as the favored nodes) for
creating any new file on the hdfs.
bq. So they will potentially "fight" – but we'll just have a perf penalty upon
recovery. Does the HDFS balancer by default run automatically (or is only
triggered manually)?
They could fight, but the hdfs balancer could in theory be tweaked to not move
blocks for certain paths. The hdfs balancer needs to be run manually.
bq. Ok, so so when we attempt to write when we compact or flush maybe we'd
check to see all N hdfs replica targets are alive? Would the tool be the only
process/thread that select data node targets? Is this the only mechanism we
have to force block replcias to different datanodes?
The tool would look at the regions, their locality information, and try to make
sure the map from regions to favored nodes is optimal. It might reassign
regions in this process (i.e., update the meta table with the new information,
that would then be propagated to the regionservers). The regionservers, like
before, would use this information (as the favored nodes) for creating any new
file on the hdfs.
> HBase based block placement in DFS
> ----------------------------------
>
> Key: HBASE-4755
> URL: https://issues.apache.org/jira/browse/HBASE-4755
> Project: HBase
> Issue Type: New Feature
> Affects Versions: 0.94.0
> Reporter: Karthik Ranganathan
> Assignee: Christopher Gist
> Priority: Critical
> Attachments: 4755-wip-1.patch
>
>
> The feature as is only useful for HBase clusters that care about data
> locality on regionservers, but this feature can also enable a lot of nice
> features down the road.
> The basic idea is as follows: instead of letting HDFS determine where to
> replicate data (r=3) by place blocks on various regions, it is better to let
> HBase do so by providing hints to HDFS through the DFS client. That way
> instead of replicating data at a blocks level, we can replicate data at a
> per-region level (each region owned by a promary, a secondary and a tertiary
> regionserver). This is better for 2 things:
> - Can make region failover faster on clusters which benefit from data affinity
> - On large clusters with random block placement policy, this helps reduce the
> probability of data loss
> The algo is as follows:
> - Each region in META will have 3 columns which are the preferred
> regionservers for that region (primary, secondary and tertiary)
> - Preferred assignment can be controlled by a config knob
> - Upon cluster start, HMaster will enter a mapping from each region to 3
> regionservers (random hash, could use current locality, etc)
> - The load balancer would assign out regions preferring region assignments to
> primary over secondary over tertiary over any other node
> - Periodically (say weekly, configurable) the HMaster would run a locality
> checked and make sure the map it has for region to regionservers is optimal.
> Down the road, this can be enhanced to control region placement in the
> following cases:
> - Mixed hardware SKU where some regionservers can hold fewer regions
> - Load balancing across tables where we dont want multiple regions of a table
> to get assigned to the same regionservers
> - Multi-tenancy, where we can restrict the assignment of the regions of some
> table to a subset of regionservers, so an abusive app cannot take down the
> whole HBase cluster.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
