[ https://issues.apache.org/jira/browse/HADOOP-10610?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14068608#comment-14068608 ]
Steve Loughran commented on HADOOP-10610: ----------------------------------------- I know it is somewhat moot as the code is now checked in, but i am disappointed that my feedback was ignored. In particular: we haven't improved code coverage through new tests, and there's still a constant embedded in the string rather than pulled out for re-use. A goal for patches should be to improve the codebase, which means through coverage and other details. for this change, in particular, what if the buffer dir is one of # the empty string, {{""}} # a list of directories that are not present/read only. I believe the allocator handles this, but tests still matter. Please can people write tests for new features, even things as minor of this. It's a change, and changes need tests > Upgrade S3n s3.fs.buffer.dir to support multi directories > --------------------------------------------------------- > > Key: HADOOP-10610 > URL: https://issues.apache.org/jira/browse/HADOOP-10610 > Project: Hadoop Common > Issue Type: Improvement > Components: fs/s3 > Affects Versions: 2.4.0 > Reporter: Ted Malaska > Assignee: Ted Malaska > Priority: Minor > Fix For: 2.6.0 > > Attachments: HADOOP-10610.patch, HADOOP_10610-2.patch, HDFS-6383.patch > > > s3.fs.buffer.dir defines the tmp folder where files will be written to before > getting sent to S3. Right now this is limited to a single folder which > causes to major issues. > 1. You need a drive with enough space to store all the tmp files at once > 2. You are limited to the IO speeds of a single drive > This solution will resolve both and has been tested to increase the S3 write > speed by 2.5x with 10 mappers on hs1. -- This message was sent by Atlassian JIRA (v6.2#6252)