[
https://issues.apache.org/jira/browse/HDFS-15493?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17175562#comment-17175562
]
Stephen O'Donnell edited comment on HDFS-15493 at 8/11/20, 1:20 PM:
--------------------------------------------------------------------
+1 on 007 patch. I will commit it later pending the CI results coming back.
I re-ran the benchmark tests on the final patch:
With Patch + Parallel loading: 202 / 203 seconds. (73612009 blocks)
No patch + parallel loading: 237 / 233 seconds
Approx 14% improvement.
With Patch (parallel load disabled): 345 / 340 seconds.
No patch (parallel load disabled): 400 / 384 seconds.
Approx 13% improvement.
The above image has significant snapshots present. [~smarthan] saw about a 20%
improvement in a large image with no snapshots.
Thanks for all the work on this [~smarthan]! I also added you to the
contributors for the HDFS project, so you should be able to assign jiras to
yourself in the future.
was (Author: sodonnell):
+1 on 007 patch. I will commit it later pending the CI results coming back.
I re-ran the benchmark tests on the final patch:
With Patch + Parallel loading: 202 / 203 seconds. (73612009 blocks)
No patch + parallel loading: 237 / 233 seconds
Approx 14% improvement.
With Patch (parallel load disabled): 345 / 340 seconds.
No patch (parallel load disabled): 400 / 384 seconds.
Approx 13% improvement.
The above image has significant snapshots present. [~smarthan] saw about a 20%
improvement in a large image with no snapshots.
Thanks for all the work on this [~smarthan]!
> Update block map and name cache in parallel while loading fsimage.
> ------------------------------------------------------------------
>
> Key: HDFS-15493
> URL: https://issues.apache.org/jira/browse/HDFS-15493
> Project: Hadoop HDFS
> Issue Type: Improvement
> Components: namenode
> Reporter: Chengwei Wang
> Assignee: Chengwei Wang
> Priority: Major
> Attachments: HDFS-15493.001.patch, HDFS-15493.002.patch,
> HDFS-15493.003.patch, HDFS-15493.004.patch, HDFS-15493.005.patch,
> HDFS-15493.006.patch, HDFS-15493.007.patch, fsimage-loading.log
>
>
> While loading INodeDirectorySection of fsimage, it will update name cache and
> block map after added inode file to inode directory. It would reduce time
> cost of fsimage loading to enable these steps run in parallel.
> In our test case, with patch HDFS-13694 and HDFS-14617, the time cost to load
> fsimage (220M files & 240M blocks) is 470s, with this patch , the time cost
> reduc to 410s.
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
---------------------------------------------------------------------
To unsubscribe, e-mail: hdfs-issues-unsubscr...@hadoop.apache.org
For additional commands, e-mail: hdfs-issues-h...@hadoop.apache.org
