On Wed, Mar 17, 2010 at 2:27 PM, Steve Franks wrote:
> Right there with you on this one. My solution was to get rid of hal :)
1. ports@ or x11@ may be a better place to discuss this.
2. If you compiled hal port with debug, then the corefile will be
useful; otherwise it's useless. If you haven't
Waste of disk space. With 2K frag size, a file of 1k will use a
minimum of 2K, wasting 1k. With 4k frag size it will waste 3k.
2010/3/17, Andrew Stesin :
> 2010/3/17 Matthew Dillon :
>> you absolutely must use a 4K fragment size (32K block size) and an
>> aligned partition when using UFS wit
2010/3/17 Matthew Dillon :
> you absolutely must use a 4K fragment size (32K block size) and an
> aligned partition when using UFS with 4K physical sector drives.
Sorry for my ignorance, but what's wrong with making this the default
setting for newfs regardless of whatever the drive is?
Tha
Right there with you on this one. My solution was to get rid of hal :)
Steve
___
freebsd-hackers@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-hackers
To unsubscribe, send any mail to "freebsd-hackers-unsubscr...@freebsd.org
: There is a sysctl, md_compress, that I turned out in my tests, but not
:working as expected.
: Why using gnop -S 4096 works well?
:
:Thiago
You are setting the sector size to 4K with gnop -S 4096 so presumably
ZFS will not do any fragmented writes smaller than that. I'm not
sure
I'll try tomorrow more zfs tests, with 1M alignment on begining of disk.
But I also remember that zfs block size its 128k, but metadata can be of
dynamic size. And we can use compressed files too.
There is a sysctl, md_compress, that I turned out in my tests, but not
working as expected.
Wh
I'll note one last thing with regards to write combining within the
drive's zone cache. Drive zone caches work very well for combing
adjacent sectors when the write zones are perfectly linear (when
the writes within each zone being tracked are perfectly linear).
But the drive z
On Wed, 17 Mar 2010 12:14:49 -0700
Oleksandr Tymoshenko wrote:
>
> On 2010-03-17, at 12:04 PM, Alex RAY wrote:
>
> > Hackers
> > help find the correct answer, in what way I can send interrupt on GPIO to
> > the userspace script.
> > I know that I can make the device, and open it in program, but
We experimented a bit with aligning fdisk (dos slices) by changing
the sector offset to 2 but I came to the conclusion that it was better
to do the alignment in disklabel / gpt / whatever higher-level
partitioner floats your boat and not mess with anything the BIOS
uses to boot
On 2010-03-17, at 12:04 PM, Alex RAY wrote:
> Hackers
> help find the correct answer, in what way I can send interrupt on GPIO to
> the userspace script.
> I know that I can make the device, and open it in program, but it is
> cumbersome.
> I know what I can run the script from the kernel, as in
Hackers
help find the correct answer, in what way I can send interrupt on GPIO to
the userspace script.
I know that I can make the device, and open it in program, but it is
cumbersome.
I know what I can run the script from the kernel, as in Linux, but I think
this is not the right way?
--
WBW
---
With UFS, I previouly did some tests (in portuguese):
Partition without begin aligned:
=>34 1953525101 ad4 GPT (932G)
34 19535251011 freebsd-ufs (932G)
newfs -U (default -f 2048 -b 16384)
# time tar xfj
/install/releases/i386/8.0-RELEASE/8.0-RELEASE/ports/ports.tgz
Thiago Damas writes:
> With UFS, I previouly did some tests (in portuguese):
Some commentary would be nice.
DES
--
Dag-Erling Smørgrav - d...@des.no
___
freebsd-hackers@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-hackers
On Wednesday 17 March 2010 18:16:09 Olivier Smedts wrote:
> Why not on geom_md ?
Thanks! After getting a "no such geom" message when I tried a couple of
commands without having created any partitions I presumed it was looking for a
DISK provider.
--
Bruce Cran
I tried with gpart, gpt scheme, begining in all block alignments possible,
34, 35, 36, 37, 38, 39 etc
Even 64
gpart create -s gpt ad4
gpart add -s 64 -t freebsd-zfs ad4
With UFS, I previouly did some tests (in portuguese):
Particao sem alinhar inicio
=>34 1953525101 ad4 GPT
2010/3/17 Bruce Cran :
> On Wednesday 17 March 2010 10:16:16 Mohacsi Janos wrote:
>> Dear FreeBSD hackers,
>> What is the situation with ATA 4K dirves in FreeBSD? Are there any
>> support for them in fdisk or disklabel?
>
> # mdconfig -a -f ddfile -S 4096
> md0
> # fdisk /dev/md0
> fdisk: cou
2010/3/17 Thiago Damas :
> I had problem with ZFS.
> With gnop -S 4096, it works well on /dev/ad{a}X.nop; but I decided to not
> use those disks.
So maybe this was not a ZFS problem but a partition misalignment problem ?
On a properly aligned partition with a physical sector size of 4KB and
a l
On Wednesday 17 March 2010 10:16:16 Mohacsi Janos wrote:
> Dear FreeBSD hackers,
> What is the situation with ATA 4K dirves in FreeBSD? Are there any
> support for them in fdisk or disklabel?
# mdconfig -a -f ddfile -S 4096
md0
# fdisk /dev/md0
fdisk: could not detect sector size
# mdconfig
I had problem with ZFS.
With gnop -S 4096, it works well on /dev/ad{a}X.nop; but I decided to not
use those disks.
2010/3/17 Mohacsi Janos
>
>
>
> On Wed, 17 Mar 2010, Dag-Erling Sm?rgrav wrote:
>
> Mohacsi Janos writes:
>>
>>> What is the situation with ATA 4K dirves in FreeBSD? Are ther
Am 16.03.2010 17:00, schrieb Pieter de Goeje:
> On Monday 15 March 2010 04:33:04 Havacci wrote:
>> How I can drop cache memory of my FreeBSD ? I search a lot about this
>> and don't find anything.
>> In Linux i usualy use this command:
>> sync; echo 3 > /proc/sys/vm/drop_caches
>
> Something compa
Slow performance,slow write speed, high delay per operation (gstat, ms/w)
above 400ms was common, with peaks of 3000ms per write). Tested on two
different machines.
Disks WD10EARS in single disk and mirror configurations.
Thiago
2010/3/17 Dag-Erling Smørgrav
> Thiago Damas writes:
> > I h
Thiago Damas writes:
> I had problem with ZFS.
What kind of problem?
DES
--
Dag-Erling Smørgrav - d...@des.no
___
freebsd-hackers@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-hackers
To unsubscribe, send any mail to "free
On Wed, 17 Mar 2010, Dag-Erling Sm?rgrav wrote:
Mohacsi Janos writes:
What is the situation with ATA 4K dirves in FreeBSD? Are there
any support for them in fdisk or disklabel?
Hmm, didn't we discuss this already? All we need to do is change the
defaults in fdisk so it rounds partition o
Hi folks!
Some time ago I was initiating the work on syscons driver ( see
http://wiki.freebsd.org/SysconsUnicodeProject ), then was too busy and my
part of the work stalled for about a year. At present I am going to continue
working on this.
One of my students, Vladislav Soldatov, is willing to c
Mohacsi Janos writes:
> What is the situation with ATA 4K dirves in FreeBSD? Are there
> any support for them in fdisk or disklabel?
Hmm, didn't we discuss this already? All we need to do is change the
defaults in fdisk so it rounds partition offsets and sizes to a multiple
of 8 sectors (or 16 f
On Wednesday 17 March 2010 03:37:12 am Andriy Gapon wrote:
> on 17/03/2010 12:16 Mohacsi Janos said the following:
> > Dear FreeBSD hackers,
> > What is the situation with ATA 4K dirves in FreeBSD? Are there
> > any support for them in fdisk or disklabel?
> > More information at:
> > http://ata
on 17/03/2010 12:16 Mohacsi Janos said the following:
> Dear FreeBSD hackers,
> What is the situation with ATA 4K dirves in FreeBSD? Are there any
> support for them in fdisk or disklabel?
> More information at:
> http://ata.wiki.kernel.org/index.php/ATA_4_KiB_sector_issues
Did you mean to say
Dear FreeBSD hackers,
What is the situation with ATA 4K dirves in FreeBSD? Are there any
support for them in fdisk or disklabel?
More information at:
http://ata.wiki.kernel.org/index.php/ATA_4_KiB_sector_issues
Janos Mohacsi
Head of HBONE+ project
Network Engineer, Deputy Director of Network P
28 matches
Mail list logo
