gzip vs. bzip2 (mainly de-)compression "benchmark"

From: Thunder from the hill (thunder@ngforever.de)
Date: Mon May 06 2002 - 12:58:42 EST


gzipping, bzip2ing a solaris disk image from my old sparc64 under alpha
(approx. 2 GiB) and relative load.
This has been done for testing purposes for those fiddling around with the
kernel compression algorithms.

# ls -l ; ls -lh
-rw-r--r-- 1 root root 2164082688 May 5 04:41 old-root
-rw-r--r-- 1 root root 2.0G May 5 04:41 old-root

# time gzip -9n old-root
9035.15user 119.40system 11:38:12elapsed 21%CPU
(0avgtext+0avgdata 0maxresident)k
0inputs+0outputs (166major+112minor)pagefaults 0swaps

# ls -l ; ls -lh
-rw-r--r-- 1 65534 65534 546148165 May 2 11:30 old-root.gz
-rw-r--r-- 1 65534 65534 521M May 2 11:30 old-root.gz

# time gzip -cd old-root.gz > /dev/null
134.11user 66.22system 10:42.91elapsed 31%CPU
(0avgtext+0avgdata 0maxresident)k
0inputs+0outputs (102major+37minor)pagefaults 0swaps

# time bzip2 -9z old-root
(Unfortunately crashed shortly after completion when I modprobed khttpd and
tried to make a connection. I can redo that if you are interested in the
results, but I tell you it took some time... user was in the 11,000 I

# ls -l ; ls -lh
-rw-r--r-- 1 root root 496780930 May 5 04:41 old-root.bz2
-rw-r--r-- 1 root root 474M May 5 04:41 old-root.bz2

# time bzip2 -cdk old-root.bz2 > /dev/null
1010.44user 33.15system 58:12.14elapsed 29%CPU
(0avgtext+0avgdata 0maxresident)k
0inputs+0outputs (117major+914minor)pagefaults 0swaps

Penguin: Linux 2.4.8 (didn't have time to compile the latest here) on an
         Alpha 300, 384 MiB RAM. Meanwhile someone was compiling Redhat
         Package Manager v4.1, later the Linux Kernel v2.4 repository.


if (errno == ENOTAVAIL)
    fprintf(stderr, "Error: Talking to Microsoft server!\n");

- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/

This archive was generated by hypermail 2b29 : Tue May 07 2002 - 22:00:27 EST