This library provides access to two compression methods, both of which generate a valid LZ4 stream:
Low memory footprint (~16 KB)
Very fast (fast scan with skipping heuristics in case the input looks incompressible)
Reasonable compression ratio (depending on the redundancy of the input)
High compression (LZ4 HC):
Medium memory footprint (~ 256 KB)
Rather slow (~10 times slower than LZ4)
A good compression ratio (depending on the size and the redundancy of the input).
The streams produced by these two compression algorithms use the same compression format; they are very fast to decompress and can be decompressed by the same decompressor instance.
LZ4 comes bundled with Hadoop. Check for the shared object(so) file, libhadoop.so, which has proper read, write, and execute rights while starting HBase. This can be done post setup too by creating a symbolic link from HBase libraries to the Hadoop native libraries.
For example, suppose your platform is Linux-amd64...