accumulo
accumulo copied to clipboard
Review HDFS file read performance enhancements
I came across https://issues.apache.org/jira/browse/HADOOP-18287 which references some new APIs:
High Performance Vectored Read API openFile ByteBufferPositionedReadable
There is also https://issues.apache.org/jira/browse/HADOOP-18287 which will enable these new features as they become available.
2.1 has been developed assuming 3.0 or later. If we do use any new APIs, we have to be careful about ensuring we don't break our minimum compatibility (which we can revisit for the next version, but I think is too late in the dev cycle for 2.1)
Agreed, this is an issue for a version after 2.1