Re: Performance comparison: Bzip2 vs Gzip

"Tristan Rhodes" <tristanr-gOv+VWruyeBS6pTDpnvEHvU/[email protected]>
Newsgroups gmane.comp.sysutils.backup.dar.general
Message-ID <[email protected]>
I agree, I had heard that Bzip2 was slower, and only provided a little better compression.  So I was surprised by the results as well.  

I backed up the entire / (root) of a basic Suse 8.2 linux system to another mounted hard drive. (I used -P proc/ -P tmp/ -P mnt/backup/  - Did I miss anything else that I should exclude?)

My hardware is a 533 MHZ Pentium III, 640 MB RAM, 7200 RPM IDE hard drives.  I tried to monitor CPU usage during backups.  When using either -z or -y (highest level) my CPU is 0% idle.  When doing a non-compressed backup, my processor stays at about 30% idle (do to a bottleneck in disk I/O). 

By the way, an uncompressed backup took about 10 minutes with a size of 1.4 GB.  Full results should be posted tonight or tomorrow, I need to test the speed of restore.

I also have a question about excluding already compressed files.  Do you use the -Z *.gz -Z *.tgz -Z *.bz2 exclusions?  (Did I miss any compressed file extensions?)

Tristan


>>> Nicola Larosa <nico-kqPTfkfBL9/[email protected]> 10/21/03 03:19AM >>>
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

> I am in the process of benchmarking Dar with different compression
> settings.  So far, it seems like Bzip2 is the better choice, both in
> compressed size AND performance.  I will post the full results when I am
> done, but here is a teaser...
>
> Creating Dar backup....
> Bzip2: 45 minutes, 590,000,000 bytes
> Gzip: 71 minutes, 610,000,000 bytes

That's peculiar, bzip2 is usually slower than gzip, at least at the default
compression levels. Did you change them? What kind of data did you compress?


> Which compression utility do other Dar users choose, and what are your
> experiences with them?

I use gzip -1 (minimum compression), good speed, decent compressed size.


- --
"Computers are in many ways becoming a bad word in our society as a
whole. They are nothing but spam, porn, and bad ergonomics."
 -- Shelley Walsh

Nicola Larosa - nico-kqPTfkfBL9/[email protected] 


-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.2 (GNU/Linux)

iD8DBQE/lPosXv0hgDImBm4RAvRoAKC5wStBSMgsi4aXbhpQu3kzJVghJACgvXsu
wBh3h5iEwA0tIJrCfGlraPE=
=+5BI
-----END PGP SIGNATURE-----



-------------------------------------------------------
This SF.net email is sponsored by OSDN developer relations
Here's your chance to show off your extensive product knowledge
We want to know what you know. Tell us and you have a chance to win $100
http://www.zoomerang.com/survey.zgi?HRPT1X3RYQNC5V4MLNSV3E54 
_______________________________________________
Dar-discussions mailing list
Dar-discussions-5NWGOfrQmneRv+LV9MX5uipxlwaOVQ5f@public.gmane.org 
https://lists.sourceforge.net/lists/listinfo/dar-discussions



-------------------------------------------------------
This SF.net email is sponsored by OSDN developer relations
Here's your chance to show off your extensive product knowledge
We want to know what you know. Tell us and you have a chance to win $100
http://www.zoomerang.com/survey.zgi?HRPT1X3RYQNC5V4MLNSV3E54
lmpx.com only provides a reader for public news (NNTP) servers. It is not affiliated with the servers or forums shown here and is not responsible for the content of articles, which is written by their respective authors.