[
https://issues.apache.org/jira/browse/HDFS-9469?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15069021#comment-15069021
]
Arpit Agarwal commented on HDFS-9469:
-------------------------------------
Hi [~anu], my comments/questions. I apologize if some of this overlaps with the
discussion you had with [~szetszwo] earlier.
# DiskBalancerCluster.java:265: Will this be called by the executor?
# DiskBalancerCluster.java:346: These threads are used only for computing plans
correct? It looks like computing a plan for a node should be quite fast once we
have collected the cluster information? Perhaps we can avoid making it
multi-threaded for now.
# GreedyPlanner.java:103: We remove the volumes from the set here and add them
back in applyStep below. Is that necessary, looks like we can skip both steps.
# GreedyPlanner.java:105: I think I need to convince myself this condition can
be hit in practice since we just recomputed idealUsed at the end of
removeSkipVolumes.
# GreedyPlanner.java:116: Same question as above.
# GreedyPlanner.java:231: Can we use MoveStep#toString instead?
# DiskBalancerTestUtil.java:36: Curious why the switch from base-2 to base-10
units as much as I dislike base-10 units.
Comments also posted at
https://github.com/arp7/hadoop/commit/06caccf07da421e6e5525ffa7099a9e06d0a88ad
> DiskBalancer : Add Planner
> ---------------------------
>
> Key: HDFS-9469
> URL: https://issues.apache.org/jira/browse/HDFS-9469
> Project: Hadoop HDFS
> Issue Type: Sub-task
> Components: balancer & mover
> Affects Versions: 2.8.0
> Reporter: Anu Engineer
> Assignee: Anu Engineer
> Attachments: HDFS-9469-HDFS-1312.001.patch,
> HDFS-9469-HDFS-1312.002.patch, HDFS-9469-HDFS-1312.003.patch
>
>
> Disk Balancer reads the cluster data and then creates a plan for the data
> moves based on the snap-shot of the data read from the nodes. This plan is
> later submitted to data nodes for execution.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
