Github user piaozhexiu commented on the pull request:
https://github.com/apache/spark/pull/8512#issuecomment-138411716
@davies I just implemented your suggestion, i.e. list each HadoopRDD using
the S3 client in parallel. The result is quite impressive!
\# of files | \# of partitions | s3 bulk listing w/ common prefix | s3 bulk
listing w/ 10 threads
---- | ------------|-----|-----
972 | 1 | 15.209 | 13.823
13646 | 24 | 13.224 | 15.31
136222 | 240 | 208.089 | 26.696
445377 | 720 | 275.298 | 48.908
Here is the graph that compares all the 4 implementations-
a: 1.5 rc3
b: hdfs fileinputformat w/ 10 threads
c: s3 bulk listing w/ common prefix
d: s3 bulk listing w/ 10 threads
---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well. If your project does not have this feature
enabled and wishes so, or if the feature is enabled but not working, please
contact infrastructure at [email protected] or file a JIRA ticket
with INFRA.
---
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
