more weird bugs with mmap-ing via NFS

Oliver Fromme olli at lurza.secnetix.de
Wed Mar 22 11:27:42 UTC 2006


Mikhail Teterin <mi+mx at aldan.algebra.com> wrote:
 > (no softupdates). It was created with `-O1 -b 65536 -f 8192' as it is intended 
 > for large files and needs no ACLs (hence no UFS1).

Those values are very suboptimal.  Whe creating a file system
for large files, you should rather decrease the inode density
(-i option).  Using "-i 262144" should work fine, or even
"-i 1048576" (larger values don't make much sense, though).

Increasing block size and fragment size like that is not a
good idea.  In fact, in earlier versions of FreeBSD there
were bugs which could lead to file system corruption when
64k block size was used.  I don't know if those bugs have
been fixed -- maybe nobody knows, because nobody uses such
large block sizes, so they aren't extensively tested.  ;-)

However, you could try setting block size and fragment size
to the same value, effectively disabling fragmentation (you
don't really need fragments when you have only large files).
For example, "-b 8192 -f 8192" should be OK.

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.

"I started using PostgreSQL around a month ago, and the feeling is
similar to the switch from Linux to FreeBSD in '96 -- 'wow!'."
        -- Oddbjorn Steffensen


More information about the freebsd-stable mailing list