bolkedebruin 7d1163b0d9 Optimize chunkedCopy for sequential writes (#5477) ...

NativeIO.chunkedCopy fsyncs its writebuffer directly and
requires an O_DIRECT RandomAccessFile. By allowing the
kernel to start writing while filling the buffer the writes
will be more constant. In addition the O_DIRECT flag is not
required anymore and this will work faster in case fadvise
is not supported on some system.

This is based on Linus' post here:
http://lkml.iu.edu/hypermail/linux/kernel/1005.2/01845.html

2018-03-14 15:15:41 -07:00

..

src

Optimize chunkedCopy for sequential writes (#5477)

2018-03-14 15:15:41 -07:00

pom.xml

Skip OS cache on Linux when pulling segments (#5421)

2018-03-08 07:54:21 -08:00