[
https://issues.apache.org/jira/browse/HDFS-9314?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15023081#comment-15023081
]
Ming Ma commented on HDFS-9314:
-------------------------------
The approach of the jira is to relax the policy regarding which replica can be
deleted so that it has more replicas to choose from to satisfy other
constraints such as storage type. Implementation wise, we modify the semantics
of {{pickupReplicaSet}} from "return the best replicas" to "return all replicas
as long as it doesn't violate the block placement policy".
For the question regarding {{BlockPlacementPolicyWithNodeGroup}}'s
{{pickupReplicaSet}}, if we apply the new semantics of {{pickupReplicaSet}},
given {{BlockPlacementPolicyWithNodeGroup}} reuses
{{BlockPlacementPolicyDefault}}'s {{verifyBlockPlacement}}, does
{{BlockPlacementPolicyWithNodeGroup}} still need to have its own
{{pickupReplicaSet}}?
> Improve BlockPlacementPolicyDefault's picking of excess replicas
> ----------------------------------------------------------------
>
> Key: HDFS-9314
> URL: https://issues.apache.org/jira/browse/HDFS-9314
> Project: Hadoop HDFS
> Issue Type: Improvement
> Reporter: Ming Ma
> Assignee: Xiao Chen
> Attachments: HDFS-9314.001.patch, HDFS-9314.002.patch,
> HDFS-9314.003.patch, HDFS-9314.004.patch, HDFS-9314.005.patch,
> HDFS-9314.006.patch
>
>
> The test case used in HDFS-9313 identified NullPointerException as well as
> the limitation of excess replica picking. If the current replicas are on
> {SSD(rack r1), DISK(rack 2), DISK(rack 3), DISK(rack 3)} and the storage
> policy changes to HOT_STORAGE_POLICY_ID, BlockPlacementPolicyDefault's won't
> be able to delete SSD replica.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
