Hi all,

I am working on a sample backup code, where i read the files per 7200 bytes and send it to server. Before sending to server, i compress each 7200 bytes using zlib compression algorithm using dictionary max length of 1.5 MB . I find zlib is slow.

Can anyone recommend me a compression technique to satisfy my above requirements.
I saw lz4 compression, but it does not support dictionary in it.

or Is there any way to improve the existing zlib algorithm ..??

./Selva..,

Higher comperssion takes more CPU cycles. You can't have it both ways.

i compress each 7200 bytes using zlib compression algorithm using dictionary max length of 1.5 MB

The compression will depend on the type of data.

Can anyone recommend me a compression technique to satisfy my above requirements.

What are the requirements? In fact, whatever they are, you'll have to trade speed for compression ratio and vice-versa, as kbw said.

In your case, you could gain speed in zlib by lowering the dictionary size?

Faster, weaker:
http://code.google.com/p/lz4/
http://www.oberhumer.com/opensource/lzo/
http://code.google.com/p/snappy/

Slower, stronger:
http://www.7-zip.org/sdk.html
http://mattmahoney.net/dc/

Why do you need to compress the data you are sending to the server? How far away is the server? Same building, or even town? How are you transmitting the data to the server? HTTPS?