Date: Mon, 17 Dec 2012 21:50:00 GMT From: "Dieter BSD" <dieterbsd@engineer.com> To: freebsd-bugs@FreeBSD.org Subject: Re: bin/163992: dumpfs(8): dumpfs -m is broken Message-ID: <201212172150.qBHLo0gE059456@freefall.freebsd.org>
next in thread | raw e-mail | index | archive | help
The following reply was made to PR bin/163992; it has been noted by GNATS.
From: "Dieter BSD" <dieterbsd@engineer.com>
To: bug-followup@freebsd.org
Cc:
Subject: Re: bin/163992: dumpfs(8): dumpfs -m is broken
Date: Mon, 17 Dec 2012 16:48:25 -0500
newfs -e 100000000 -b 65536 -f 8192 -g 67108864 -h 16 -i 67108864
-U -o space -L ${DISK_NUM}data /dev/${DISK_PORT}
newfs says:
density reduced from 67108864 to 14860288
/dev/ada12: 2861588.5MB (5860533168 sectors) block size 65536,
fragment size 8192
using 789 cylinder groups of 3628.00MB, 58048 blks, 256 inodes.
with soft updates
super-block backups (for fsck -b #) at:
256, 7430400, 14860544, 22290688, 29720832, [ ... ]
The new dumpfs says:
newfs -L 71data -O 2 -U -a 2 -b 65536 -d 65536 -e 100000000 -f 8192
-g 67108864 -h 16 -i 14868480 -m 8 -o space -s 5860533168 /dev/ufs/71data
A definite improvement, Thank you.
However...
The -i argument is off by 8K.
At least it is off in the desired direction. :-)
The -a 2 is troubling since the default is supposed to be 16 [1].
And is a rotational delay still desirable with modern disks?
(large on-disk buffers, variable number of sectors/track, ...)
I've posted a query about this to the fs and performance lists.
-d is supposed to default to 16 times the file system blocksize [1].
[1] According to the newfs man page.
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?201212172150.qBHLo0gE059456>