[
https://issues.apache.org/jira/browse/HDFS-11565?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15956200#comment-15956200
]
Andrew Wang commented on HDFS-11565:
------------------------------------
One optimization made by this patch is using ErasureCodingPolicies deduplicates
the built-in ErasureCodingPolicy objects in memory. We've seen significant
memory taken by FileStatus fields when profiling applications like Hive.
Later, we should also consider deduplicating the pluggable EC policies. Should
be relatively straightforward to do.
> Use compact identifiers for built-in ECPolicies in HdfsFileStatus
> -----------------------------------------------------------------
>
> Key: HDFS-11565
> URL: https://issues.apache.org/jira/browse/HDFS-11565
> Project: Hadoop HDFS
> Issue Type: Improvement
> Components: erasure-coding
> Affects Versions: 3.0.0-alpha2
> Reporter: Andrew Wang
> Assignee: Andrew Wang
> Priority: Blocker
> Labels: hdfs-ec-3.0-must-do
> Attachments: HDFS-11565.001.patch
>
>
> Discussed briefly on HDFS-7337 with Kai Zheng. Quoting our convo:
> {quote}
> From looking at the protos, one other question I had is about the overhead of
> these protos when using the hardcoded policies. There are a bunch of strings
> and ints, which can be kind of heavy since they're added to each
> HdfsFileStatus. Should we make the built-in ones identified by purely an ID,
> with these fully specified protos used for the pluggable policies?
> {quote}
> {quote}
> Sounds like this could be considered separately because, either built-in
> policies or plugged-in polices, the full meta info is maintained either by
> the codes or in the fsimage persisted, so identifying them by purely an ID
> should works fine. If agree, we could refactor the codes you mentioned above
> separately.
> {quote}
--
This message was sent by Atlassian JIRA
(v6.3.15#6346)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
