"tar -c|gzip" faster than "tar -cz"?!?

[Oliver Fromme]

Mike Meyer wrote:
 > Not necessarily a known problem, but not a surprise. I'm not sure
 > about the size issue - it's not clear what compression level it's
 > running at. The real time difference is expected. tar uses libarchive,
 > which does the compression in the process. So while piping tar's
 > output to gzip will let gzip compress data while tar is waiting on
 > disk I/O, having tar compress things for you means that doesn't
 > happen.

There was no disk I/O involved in my tests.  All data was
cached in RAM.  You can also see from the my numbers that
the "user time" was almost the same as the "real time".

 > Since they use different code - with different licenses - some
 > difference is expected.

Different code?  tar/libarchive uses libz, and I thought
that gzip also uses libz, but I could be wrong.

If gzip uses its own code instead of libz, that would
explain the results of my test, of course.  So it seems
that gzip is 30% faster than libz ...  quite significant,
I think.

It seems I won't use tar's z option anymore.  :-)

Best regards
   Oliver

-- 
Oliver Fromme,  secnetix GmbH & Co. KG, Marktplatz 29, 85567 Grafing
Dienstleistungen mit Schwerpunkt FreeBSD: http://www.secnetix.de/bsd
Any opinions expressed in this message may be personal to the author
and may not necessarily reflect the opinions of secnetix in any way.

"To this day, many C programmers believe that 'strong typing'
just means pounding extra hard on the keyboard."
        -- Peter van der Linden