headers
Pawel Jakub Dawidek | 1 Jun 2007 06:26
Picon
Favicon

Re: Difference between setting geli(8) key when attached or detached

On Fri, Jun 01, 2007 at 12:44:27AM +0200, Jeremie Le Hen wrote:
> Hi Pawel,
> 
> I dare to contact you because I'm studying GELI's code and I found
> a piece of code I'm not sure to understand, although I've read phk's
> GEOM tutorial thoroughly.
> 
> >From what I've undertood (please, correct me if I'm wrong), a
> "spoiled" event is ``posted when a provider gets a non-zero access
> count.  All attached providers, except the guilty party, are
> notified.''

s/non-zero access count/non-zero write access count/

Here is the thing. When your class makes decisions based on provider's
on-disk metadata, you want to receive spoil event and self-destruct,
because open for write means that someone may modify your metadata.
Then, on last write close, taste event is send and your class can read
eventually modified metadata once again.

> geli(8)'s "setkey" command uses two different code paths, depending
> on whether the provider is attached or not.  If is it attached,
> it seems to use the GEOM kernel part to update the key while
> if it is detached it writes it directly from userland.
> 
> My thought is that the provider being modified is not notified
> by the GEOM framework.  Am I right?

GELI doesn't do autoconfiguration. The only place when GELI uses taste
event is before root file system is mounted, so it can ask for a
(Continue reading)

Permalink | Reply | 501 comments |
headers
Mike Meyer | 3 Jun 2007 07:49
X-Face
Gravatar

WARNING: Expected rawoffset 0, found 63

Hello,

I'm changing the way I'm mirror my system disk on an amd64 system
running 6-STABLE (from late april). In paticular, I want to mirror the
root partition instead of the entire disk, and have this setup for
that:

bhuda% gmirror list gm0s1a
Geom name: gm0s1a
State: COMPLETE
Components: 2
Balance: split
Slice: 4096
Flags: NONE
GenID: 0
SyncID: 1
ID: 2799177281
Providers:
1. Name: mirror/gm0s1a
   Mediasize: 1073741312 (1.0G)
   Sectorsize: 512
   Mode: r1w1e1
Consumers:
1. Name: ad2s1a
   Mediasize: 1073741824 (1.0G)
   Sectorsize: 512
   Mode: r1w1e1
   State: ACTIVE
   Priority: 0
   Flags: NONE
(Continue reading)

Permalink | Reply | 0 comments |
headers
FreeBSD bugmaster | 4 Jun 2007 13:08
Picon
Favicon

Current problem reports assigned to you

Current FreeBSD problem reports
Critical problems
Serious problems

S Tracker      Resp.      Description
--------------------------------------------------------------------------------
o kern/73177   geom       kldload geom_* causes panic due to memory exhaustion
o kern/76538   geom       [gbde] nfs-write on gbde partition stalls and continue
o kern/83464   geom       [geom] [patch] Unhandled malloc failures within libgeo
o kern/84556   geom       [geom] GBDE-encrypted swap causes panic at shutdown
o kern/87544   geom       [gbde] mmaping large files on a gbde filesystem deadlo
o kern/89102   geom       [geom_vfs] [panic] panic when forced unmount FS from u
o bin/90093    geom       fdisk(8) incapable of altering in-core geometry
o kern/90582   geom       [geom_mirror] [panic] Restore cause panic string (ffs_
o kern/98034   geom       [geom] dereference of NULL pointer in acd_geom_detach 
o kern/104389  geom       [geom] [patch] sys/geom/geom_dump.c doesn't encode XML

10 problems total.

Non-critical problems

S Tracker      Resp.      Description
--------------------------------------------------------------------------------
o bin/78131    geom       gbde "destroy" not working.
o kern/79251   geom       [2TB] newfs fails on 2.6TB gbde device
o kern/94632   geom       [geom] Kernel output resets input while GELI asks for 
f kern/105390  geom       [geli] filesystem on a md backed by sparse file with s
o kern/107707  geom       [geom] [patch] add new class geom_xbox360 to slice up 
p bin/110705   geom       gmirror control utility does not exit with correct exi
(Continue reading)

Permalink | Reply | 0 comments |
headers
Francisco Reyes | 4 Jun 2007 13:40

geom_gate and geom_mirror status?

Anyone has any experience with geom_gate, geom_mirror features in an ISP 
style setup?

At work we are testing them and so far it looks promissing, but we are 
wondering how it behaves under heavy load. Would like to
use it for replicating Maildir directories at an ISP in real time.

Originally we were looking for DRBD in FreeBSD (http://www.drbd.org) and 
searching for that we bumped into references to geom_gate/geom_mirror.
Permalink | Reply | 3 comments |
headers
Arne "Wörner" | 4 Jun 2007 14:30
Picon
Favicon

Re: geom_gate and geom_mirror status?

--- Francisco Reyes <lists <at> stringsutils.com> wrote:
> Anyone has any experience with geom_gate, geom_mirror features in an ISP 
> style setup?
>
I dont...

> At work we are testing them and so far it looks promissing, but we are 
> wondering how it behaves under heavy load. Would like to
> use it for replicating Maildir directories at an ISP in real time.
>
The problem would be, that UFS (I assume that u want to use UFS on top of
gmirror) does not allow multiple RW mounts...

Maybe u would like afs/dfs or coda or mysql (i dont know if there r mail
exchangers, that can use mysql; maybe qmail?) better?

Or maybe u would like this self-made construction better:
Whenever an email comes in from the outside (so not from the other hot box), u
forward it via some specially crafted /etc/aliases file to the other hot box.
Then u would need something, that pop3-s both boxes simultaneously somehow
(maybe via a journal, that is sent to the other hot box every few seconds, and
that is shortened if the changes r acknowledged)... And after a crash the box
would be updated automatically via the mail queue (if the down time was not too
long)...

-Arne

      ___________________________________________________________________________________
You snooze, you lose. Get messages ASAP with AutoCheck
in the all-new Yahoo! Mail Beta.
(Continue reading)

Permalink | Reply | 0 comments |
headers
Tomas Zvala | 4 Jun 2007 15:22
Picon

Re: geom_gate and geom_mirror status?

Francisco Reyes wrote:
> Anyone has any experience with geom_gate, geom_mirror features in an 
> ISP style setup?
>
> At work we are testing them and so far it looks promissing, but we are 
> wondering how it behaves under heavy load. Would like to
> use it for replicating Maildir directories at an ISP in real time.
>
> Originally we were looking for DRBD in FreeBSD (http://www.drbd.org) 
> and searching for that we bumped into references to 
> geom_gate/geom_mirror.
>
About a year ago I tried setting up similar solution with FBSD and 
ggate/gmirror but even with generic kernel it wasn't stable enough for 
production usage. I tried to tweak around a bit but never succeeded in 
having it rock solid. Another problem is the need to do full resync 
after disconnect event (and with 900G disk it takes a looooong time even 
on 1GBps connection). So our mailserver runs gentoo linux with drbd for 
about 10 months now and works perfectly.

(just for clarification, we use the other box just as an offline backup 
-> fs is not mounted and we just have a manual failover in the event of 
master server failure)

Tomas
Permalink | Reply | 1 comment |
headers
Francisco Reyes | 4 Jun 2007 15:59

Re: geom_gate and geom_mirror status?

Tomas Zvala writes:

> About a year ago I tried setting up similar solution with FBSD and 
> ggate/gmirror but even with generic kernel it wasn't stable enough for 
> production usage.

Thanks much for the feedback.
We have set it up in two test machines and it hung up with some basic 
testing. This was with 6.2 Release. 

Doesn't look very promissing so far.

> having it rock solid. Another problem is the need to do full resync 
> after disconnect event (and with 900G disk it takes a looooong time even 
> on 1GBps connection)

We are going to have 50GB to 200GB. Unlikely any more than that.

> So our mailserver runs gentoo linux with drbd for 
> about 10 months now and works perfectly.

I wonder if anyone has ever tried to port drbd to FreeBSD.

> -> fs is not mounted and we just have a manual failover in the event of 
> master server failure)

Take a look at ucarp. A co-worker is doing the testing of our high 
availability setup. He said that ucarp was pretty straight forward to setup.
Permalink | Reply | 0 comments |
headers
Fluffles | 4 Jun 2007 16:49

GEOM an acryonym?

Hello list,

We all know GEOM, but where GEOM got it's name from is not very well 
documented. I found two sources claiming it to be an acronym for 
"Geometry" and would like to verify this. The sources are:

http://people.freebsd.org/~phk/Geom/
http://acronyms.thefreedictionary.com/GEOM

Both leave something to be desired: the first source is from it's author 
but does not indicate if GEOM is truely derived from Geometry. The 
second does not state if the name is to be seen in FreeBSD context. I'm 
writing the GEOM-article on english wikipedia and would like this 
clarified, thanks!

- Veronica
Permalink | Reply | 2 comments |
headers
Poul-Henning Kamp | 4 Jun 2007 17:14
Picon
Favicon

Re: GEOM an acryonym?

In message <4664266D.8090704 <at> fluffles.net>, Fluffles writes:
>Hello list,
>
>We all know GEOM, but where GEOM got it's name from is not very well 
>documented. I found two sources claiming it to be an acronym for 
>"Geometry" and would like to verify this. The sources are:
>
>http://people.freebsd.org/~phk/Geom/
>http://acronyms.thefreedictionary.com/GEOM
>
>Both leave something to be desired: the first source is from it's author 
>but does not indicate if GEOM is truely derived from Geometry.

It is.  Truly.

--

-- 
Poul-Henning Kamp       | UNIX since Zilog Zeus 3.20
phk <at> FreeBSD.ORG         | TCP/IP since RFC 956
FreeBSD committer       | BSD since 4.3-tahoe    
Never attribute to malice what can adequately be explained by incompetence.
Permalink | Reply | 0 comments |
headers
RW | 4 Jun 2007 17:06
Picon

Re: GEOM an acryonym?

On Mon, 04 Jun 2007 16:49:17 +0200
Fluffles <etc <at> fluffles.net> wrote:

> Hello list,
> 
> We all know GEOM, but where GEOM got it's name from is not very well 
> documented. I found two sources claiming it to be an acronym for 
> "Geometry" and would like to verify this.

By definition GEOM is not an acronym for Geometry.
Permalink | Reply | 0 comments |

Gmane