gjournal: what is it good for?

[David N]

On 21 April 2010 12:48, Andrew Reilly <areilly at bigpond.net.au> wrote:
> Hi David,
>
> On Wed, Apr 21, 2010 at 12:20:46PM +1000, David N wrote:
>> Gjournal on my systems are pretty quick on startup after a power outtage.
>
> Great to hear!
>
>> What kind of disks are you using? Or what hardware are you using?
>
> Several: main /usr is on a gjournal on top of a gmirror over a
> pair of Samsung 1TB 3.5" SATA drives, but I have other gjournals
> on a 750G WD SATA, a 1.5T Seagate and another 1TB WD MyBook
> firewire unit.  There is brokenness in the firewire connection
> that results in me always coming up manually through single-user
> mode, at the moment.  In single user mode pilot error is
> sufficient to account for the problems that I was having with
> mount vs fsck of the gjournalled drives, I'm fairly sure.
>
> The firewire issue is strange: at boot-up the fw stack (this
> is 9-current) doesn't reset the bus well enough or for long
> enough for the drive to be fully recognised and show up in /dev,
> but a post-boot "fwcontrol -r" does the job nicely.  It's a bit
> inconvenient that fwcontrol is in /usr/sbin, rather than /sbin,
> so it's not available until the rest of the mounts are complete.
> USB "works" but is sufficiently slower that I'm prepared to
> hand-hold the boot process in order to have my backups finish in
> reasonable time.
>
>> Might be the disks are ignoring the BIO_FLUSH.
>
> Not sure: how could I tell?
>
> Thanks for the support and suggestions.
>
> Cheers,
>
> --
> Andrew
>

Wow, thats a setup.
I have a few more questions.

Your first email, you mentioned gjournal overflow panics. Have you fixed that?
I see you are gmirroring the slices, when you did the gmirror +
gjournal slice, did you check the bsdlabel? sometimes it doesn't
report the right block size, and the disk + journal overlap. I had
this happen to me on my first setup which resulted in the overflow
panics.

Its not as easy as a
gmirror label ...
gjournal label ...
You gotta check the bsdlabel each time to make sure the c slice and
additional slices are the correct size.
If you do decide to do it again, gpt made it really easy.

Did you use newfs -J to format the slices/journal?

Regards
David N