Andriy Gapon wrote this message on Tue, Feb 15, 2022 at 08:44 +0200:
> On 15/02/2022 01:17, John-Mark Gurney wrote:
> > Andriy Gapon wrote this message on Mon, Feb 07, 2022 at 14:15 +0200:
> >> I've got a problem where fsck behaves differently from my expectations.
> >
t now, the ggate protocol (from what I remember) doesn't have
a way to know when the remote kernel has received notification that an
IO is complete.
I guess this situation isn't any worse than it is right now w/o passing
the _ORDERED flag down though.
> I've done some exp
. I'm not
sure how to do this though, as tunefs and fsdb don't seem to have options
to do this, and likely you'll want to update all the superblocks w/ this
new value.
--
John-Mark Gurney Voice: +1 415 225 5579
"All that I will do, has been done, All that I have, has not."
Johannes Totz wrote this message on Sun, Sep 19, 2021 at 17:27 +0100:
> On 14/09/2021 22:21, John-Mark Gurney wrote:
> > Johannes Totz wrote this message on Mon, Sep 13, 2021 at 02:00 +0100:
> >> On 09/09/2021 23:33, John-Mark Gurney wrote:
> >>> Johannes Totz via f
Johannes Totz wrote this message on Mon, Sep 13, 2021 at 02:00 +0100:
> On 09/09/2021 23:33, John-Mark Gurney wrote:
> > Johannes Totz via freebsd-geom wrote this message on Thu, Sep 02, 2021 at
> > 21:55 +0100:
> >> (looks like gmane swallowed my 1st message, trying ag
d security)
allowing for better security..
The ssh branch does have an rc.d script for that. I'll look at which one
should be used, and adapting the man page that I've written for that one
as well.
Thanks for the help/work!
--
John-Mark Gurney Voice: +1 41
will not handle
that gracefully... (Just realized that'd be a useful mode for geli
to have, to return zeros for failed authentication.)
--
John-Mark Gurney Voice: +1 415 225 5579
"All that I will do, has been done, All that I have, has not."
>> going through that thread. What can I do to improve this situation?
> > >> Would
> > >>> it make sense to enable direct dispatch for geli? That would hurt
> > >>> single-threaded performance, but probably improve performance for
> > highly
>
g from a USB drive which can
contain the key file:
https://forums.freebsd.org/threads/zfs-boot-from-usb.45880/
I don't think zfsboot (which is needed for ZFS root booting) can handle
key files, because it needs to get the key file from somewhere, and it
is a very small binary, and so does not have the
ich I think only uses a single write thread, so
> maybe reducing that would help. But theres still a very large
> discrepancy between what all of my other jetricsare showing me and gstat.
>
> Te underlying discs are SSD's by t
u can run top -S to see
how much CPU the geli threads are using.
Also, how many eli volumes do you have? In my case, I have 13... In
order to reduce the load on the scheduler, I have:
kern.geom.eli.threads="1"
in /boot/loader.conf in order to prevent creating 6 threads (I have a
6
.
UFS "knew" about the bsdlabel, and always skipped the first 16 sectors
to not overwrite them.
If you tried this on pre-geom machines, you would have corrupted your
bsdlabel.
--
John-Mark Gurney Voice: +1 415 225 5579
"All that I will do, ha
geom" test is still using geom, just not w/ the graid3
module.
--
John-Mark Gurney Voice: +1 415 225 5579
"All that I will do, has been done, All that I have, has not."
___
freebsd-geom@freebsd.org mail
BOOT flag, which makes no sense on
> a onetime anyway, so nobody ever bothered trying. But with -[tT],
> somebody might try. Evidence: I did, and found the panic :)
Oh, it looks like we have a flag to show if a provider can delete,
see:
svnweb.freebsd.org/changeset/base/r279913
a sector has been deleted, it is possible that the data may
still be readable depending upon how the underlying provider implements
.Dv BIO_DELETE .
BIO_DELETE should be used w/ .Dv.
The date is not bumped... Even though it will be incorrect, it's useful
to include to remind people that i
any encryption at
> > all?
>
> 100% is not available yet introduced GELI keys / mounted drive.
> AES-XTS is good but too slow.
You do realize that AES-XTS can get >900MB/sec on md mounted disks when
AES-NI is available? and that is single cpu... Make sure you're
runnin
; data[i] ^= ((uint8_t*)ctx->tweak)[i];
> }
>
> if (do_encrypt)
> rijndael_encrypt(&ctx->key1, data, data);
> else
> rijndael_decrypt(&ctx->key1, data, data);
>
> if (ALIGNED_POINTER(b
= 102841051 bytes/sec
> AES-XTS-256 = 80813644 bytes/sec
Is this from a different patch or what? Can you talk more about this?
Thanks.
--
John-Mark Gurney Voice: +1 415 225 5579
"All that I will do, has been done, All that I have, has not.
ive, returned a read error and
kicked the drive out of the mirror instead of fixing it (which I would
have preferred)...
Even having a simple mode that upon read error, would read from the other
drive and write back would be good... We'd need to have a way to say
that this drive is FAILING, bu
John-Mark Gurney wrote this message on Fri, Oct 17, 2014 at 09:58 -0700:
> Sourish Mazumder wrote this message on Fri, Oct 17, 2014 at 17:34 +0530:
> > I am planning to use geom gate network for accessing remote disks. I set up
> > geom gate as per the freebsd handbook. I am us
, you'll be limited to 128k / rtt of bandwidth.
Second is that ggate isn't issueing multiple IOs at a time. This means
that any NCQ or tagging isn't able to be used, where as when running
natively they can be used...
--
John-Mark Gurney
Karl Pielorz wrote this message on Tue, Sep 30, 2014 at 22:20 +0100:
> --On 30 September 2014 06:54:45 -0700 John-Mark Gurney
> wrote:
>
> >gpt doesn't let you do gpt in gpt.. which explains why you can't
> >create the second level of gpt...
> >
> >Wh
Andrey V. Elsukov wrote this message on Tue, Sep 30, 2014 at 18:22 +0400:
> On 30.09.2014 17:54, John-Mark Gurney wrote:
> >> If I repeat the above steps without the initial GPT partition (i.e.
> >> against
> >> /dev/da0) - I end up with '/dev/da0.e
partition (i.e. against
> /dev/da0) - I end up with '/dev/da0.eli' - which I can then GPT fine etc.
gpt doesn't let you do gpt in gpt.. which explains why you can't
create the second level of gpt...
Why do you want/need to partition again? You could pos
> >case. Now I can pass multiple devices to the kernel's geom handler
> >and they get passed to the kernel ctl handler.
> >
> >The following patch should, I believe, work -- but it doesn't. The
> >first disk attaches but the second one that was
> I do not have a ahci_load=YES in /boot/loader.conf (this is on FreeBSD
> 9.2-p6) and I don't know if that's still necessary or not. Searching the
> internet turned up mostly rather old (2010,2011) results.
>
>
> Am 2014-06-03 22:48, schrieb John-Mark Gurney:
>
ther,
so again, won't have work to do until another IO is submitted...
Try sending more IO at it, like doing 4 or more dd read's such that
the between the latency of one IO, there is other IO to server...
Also, make sure that you're using NCQ where the OS can submit multiple
IO
Pawel Jakub Dawidek wrote this message on Mon, Jun 02, 2014 at 22:26 +0200:
> On Mon, Jun 02, 2014 at 11:01:08AM -0700, John-Mark Gurney wrote:
> > Michael W. Lucas wrote this message on Mon, Jun 02, 2014 at 11:36 -0400:
> > > On Mon, Jun 02, 2014 at 10:45:52AM -0400, Ryan Stone
ot;bind" the two together so
that when one opens, the other doesn't disappear... The problem is
that geom views them as two separate disks when in fact they are the
same... someone who knows geom well should think about how to solve
this problem, as diskid isn'
e the glabel man page to document the diskid feature that
you added.
Thanks.
--
John-Mark Gurney Voice: +1 415 225 5579
"All that I will do, has been done, All that I have, has not."
___
freebsd-geom@freeb
Ronald F. Guilmette wrote this message on Tue, May 20, 2014 at 13:17 -0700:
>
> In message <20140520185935.gs43...@funkthat.com>,
> John-Mark Gurney wrote:
>
> >Ronald F. Guilmette wrote this message on Sat, May 17, 2014 at 17:00 -0700:
> >> I had forgotten all
Ronald F. Guilmette wrote this message on Sat, May 17, 2014 at 17:00 -0700:
> In message <20140517162513.gg43...@funkthat.com>,
> John-Mark Gurney wrote:
>
> >Ronald F. Guilmette wrote this message on Fri, May 16, 2014 at 12:38 -0700:
> >>
> >> In messa
Ronald F. Guilmette wrote this message on Fri, May 16, 2014 at 12:38 -0700:
>
> In message <20140516130346.gb43...@funkthat.com>,
> John-Mark Gurney wrote:
>
> >> The result of the above operations is as follows:
> >>
> >> May 15 21:53:33 segfau
oblem...
What do "diskinfo /dev/ada2" and "camcontrol identify ada2" return? If
it really does return that the disk is only 31MB, we need to track this
down, and this is why we're failing the integrity check of the MBR..
Thanks.
--
John-Mark Gurney
/resources...
> > gpart backup produces plain text that can be manually created. I'd
> > still rather let gpart deal with the disk directly by using gpart create
> > and add commands.
> I'm under the assumption that 'gpart create' will do more than
> &
D8C.7000605
--
John-Mark Gurney Voice: +1 415 225 5579
"All that I will do, has been done, All that I have, has not."
___
freebsd-geom@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/free
Konstantin Belousov wrote this message on Thu, Dec 05, 2013 at 10:25 +0200:
> On Wed, Dec 04, 2013 at 06:07:43PM -0800, John-Mark Gurney wrote:
> > Konstantin Belousov wrote this message on Wed, Dec 04, 2013 at 18:20 +0200:
> > > On Wed, Dec 04, 2013 at 01:59:52PM +0100,
d really uses pbuf, and since geom_uzip knows nothing
> about unmapped bio, the assertion must not trigger.
Ummm... what's the point of MAXPHYS if we allow IO larger than it?
--
John-Mark Gurney Voice: +1 415 225 5579
"All that I will do, has been do
n gpt partitions(and don't know
> why!). After all, I have decided to back to mbr because have not found any
> particular reason to stay on gpt scheme and everything is OK with mbr
> scheme.
>
> Thank you for your useful guides :)
With out actual information, I can't h
h bagade wrote this message on Sat, Nov 23, 2013 at 09:14 +0330:
> On Thu, Nov 21, 2013 at 11:36 PM, John-Mark Gurney wrote:
>
> > h bagade wrote this message on Thu, Nov 21, 2013 at 12:24 +0330:
> > > On Thu, Nov 21, 2013 at 11:37 AM, John-Mark Gurney
> > wrote:
&
> >>
> > >>
> > >> http://pastebin.com/x5MinNQz
> > >>
> > >> why FreeBSD doesn't find microdrive when it has partitions or partition
> > >> schemas ?
> > >>
> > >> Beri
> > >>
> >
h bagade wrote this message on Thu, Nov 21, 2013 at 12:24 +0330:
> On Thu, Nov 21, 2013 at 11:37 AM, John-Mark Gurney wrote:
>
> > h bagade wrote this message on Thu, Nov 21, 2013 at 11:16 +0330:
> > > I've tried to encrypt my disk in freebsd 9.2 based on the follo
now, don't be disappointed w/ the performance of geli
in 9.2... I plan on back porting the changes I've made to HEAD back
to 9 at some point, and then you'll finally get decent performance..
Most of my changes will be in 10.0-R, we'll see if the my l
ks, and
if it would work well for geli devices that ask for passphrases on
boot..
P.S. Apparently a lot more people are using geli that I suspected.
I'm happy that this is happening, but we have a lot of work to make
it more usable.
Thanks!
--
John-Mark Gurney
er write
could come in and complete on the second drive before the first IO
completes breaking the assumptions that UFS+S and ZFS about how/when
IO gets committed to the disk...
Also, depending upon the write load, it could be the HD is faster than
the SSD if there was a l
wanted to fix.
Though I'm volunteering to help, don't forget that most patches should
still be reviewed publicly, so you'll want to keep working on the -geom
mailing list.
Thanks for your interest and work.
--
John-Mark Gurney Voice: +1 415 225 5579
s
are similar...
> > would like to read key from a file for partition, which holds root file
> > system (so you need the key after the kernel is loaded, but before root
> > file system is mounted) then no, it is not currently possible. Key can
> :-(
--
John-Mark Gurney
Pawel Jakub Dawidek wrote this message on Sun, Sep 23, 2012 at 08:43 +0200:
> On Sat, Sep 22, 2012 at 09:48:28PM -0700, John-Mark Gurney wrote:
> > Pawel Jakub Dawidek wrote this message on Sat, Sep 22, 2012 at 18:20 +0200:
> > > On Tue, Sep 18, 2012 at 09:04:30PM -0700, John
Pawel Jakub Dawidek wrote this message on Sat, Sep 22, 2012 at 18:20 +0200:
> On Tue, Sep 18, 2012 at 09:04:30PM -0700, John-Mark Gurney wrote:
> > I was looking at geli and I'm not sure if it's implementing BIO_FLUSH
> > and/or BIO_ORDERED properly...
> >
&g
n that a _FLUSH is immediately dispatched,
even when there may be _WRITEs that haven't been finished encrypting and
sent down to the disk to get _FLUSHed...
Any comments about this?
Thanks.
--
John-Mark Gurney Voice: +1 415 225 5579
"All that I will do, ha
the man page if someone comes up w/
a good list of points. Are the ones I listed above enough?
--
John-Mark Gurney Voice: +1 415 225 5579
"All that I will do, has been done, All that I have, has not."
___
fr
n...
> But such situations, IMHO, are not very often ones.
Won't gmirror still report COMPLETE after a gmirror remove? So the
script can look at the gmirror device, and see that it is still
complete even though one of the providers were dropped and assume
it was an administrative comm
52 matches
Mail list logo
