[
https://issues.apache.org/jira/browse/HADOOP-15547?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16516518#comment-16516518
]
Steve Loughran commented on HADOOP-15547:
-----------------------------------------
core sounds good; not in a position to test this right now (travelling).
If you are tuning list, make sure that listStatus() is covered too, to see if
you can go from a directory walk to a flatter scan & incremental results...HDFS
uses it to avoid marshalling massive arrays; for S3A we can switch from a slow
treewalk to a files/5000 LIST calls, irrespective of depth. And no, no need to
guarantee consistency of the list.
I don't we have any list scale tests for use across stores. There is
{{ITestS3ADirectoryPerformance}} to use as a source...we can't just pull that
up as its counting internal FS metrics and asserting on their values.
> WASB: listStatus performance
> ----------------------------
>
> Key: HADOOP-15547
> URL: https://issues.apache.org/jira/browse/HADOOP-15547
> Project: Hadoop Common
> Issue Type: Bug
> Components: fs/azure
> Affects Versions: 2.9.1, 3.0.2
> Reporter: Thomas Marquardt
> Assignee: Thomas Marquardt
> Priority: Major
> Attachments: HADOOP-15547.001.patch, HADOOP-15547.002.patch,
> HADOOP-15547.003.patch
>
>
> The WASB implementation of Filesystem.listStatus is very slow due to O(n!)
> algorithm to remove duplicates and uses too much memory due to the extra
> conversion from BlobListItem to FileMetadata to FileStatus. It takes over 30
> minutes to list 700,000 files.
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
