[
https://issues.apache.org/jira/browse/HDFS-6268?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Jason Lowe updated HDFS-6268:
-----------------------------
Fix Version/s: 2.5.0
We also ran into the massive skew issue during localization that
[~ashwinshankar77] encountered. The result previously was not deterministic
since it would swap in a random node to the first position if it wasn't local
or rack-local, but now it always sends all off-rack requests to the same node.
Localization is a pretty common process, so I'm not sure sending most of the
nodes to a single datanode is a good default. Filed HDFS-6840 to discuss this
further.
> Better sorting in NetworkTopology#pseudoSortByDistance when no local node is
> found
> ----------------------------------------------------------------------------------
>
> Key: HDFS-6268
> URL: https://issues.apache.org/jira/browse/HDFS-6268
> Project: Hadoop HDFS
> Issue Type: Improvement
> Affects Versions: 2.4.0
> Reporter: Andrew Wang
> Assignee: Andrew Wang
> Priority: Minor
> Fix For: 3.0.0, 2.5.0
>
> Attachments: hdfs-6268-1.patch, hdfs-6268-2.patch, hdfs-6268-3.patch,
> hdfs-6268-4.patch, hdfs-6268-5.patch, hdfs-6268-branch-2.001.patch
>
>
> In NetworkTopology#pseudoSortByDistance, if no local node is found, it will
> always place the first rack local node in the list in front.
> This became an issue when a dataset was loaded from a single datanode. This
> datanode ended up being the first replica for all the blocks in the dataset.
> When running an Impala query, the non-local reads when reading past a block
> boundary were all hitting this node, meaning massive load skew.
--
This message was sent by Atlassian JIRA
(v6.2#6252)
