buh...@lothlorien.nfbcal.org (Brian Buhrow) writes:
> Hello. Sorry I wasn't clear. The patch I sent to the mailing list
>does all the things you outline causing the vnd disk driver to now behave
>comparably to other disk drivers. I take it you're in favor of committing
>the patch.
I'm in
Hello. Sorry I wasn't clear. The patch I sent to the mailing list
does all the things you outline causing the vnd disk driver to now behave
comparably to other disk drivers. I take it you're in favor of committing
the patch.
-thanks
-Brian
On Jun 7, 5:47am, Michael van Elst wrote:
} Su
buh...@lothlorien.nfbcal.org (Brian Buhrow) writes:
Hello Brian,
> hello. In researching a recent bug under NetBSD-5, I discovered the
>desire to make wedges work with vnd disks. While the ability for these
>devices to deal with wedge ioctls was added by riz some time ago, the
>ability to
hello. In researching a recent bug under NetBSD-5, I discovered the
desire to make wedges work with vnd disks. While the ability for these
devices to deal with wedge ioctls was added by riz some time ago, the
ability to have wedges autodiscovered when the vnd is configured was not.
The be
On Wed, Jun 06, 2012 at 06:49:32PM +0200, Edgar Fu? wrote:
>
> Are stripes units larger than MAXPHYS impossible?
Yes.
--
Thor Lancelot Simont...@panix.com
"We cannot usually in social life pursue a single value or a single moral
aim, untroubled by
> Of course, this is all still theoretical -- the best thing is to
> experiment with different RAID settings and real workloads to see what
> works best for the particular applications...
So I've bitten the bullet and conducted experiments.
I hope the results are useful to others, too.
The setup
I tried to create a RAID 5 with 512 sectPerSU and upon raidctl -i, I get
mpt0: error 22 loading DMA map
sd6(mpt0:0:7:0): generic HBA error
raid1: IO Error. Marking /dev/sd6a as failed.
sd5(mpt0:0:5:0): generic HBA error
sd4(mpt0:0:4:0): generic HBA error
Hello
For now the kernel has a global maxvndes limit, and starts reclaiming
vnodes once the limit is reached. That means unused vnodes will remain
for a while in the kernel. This may be a problem for userland filesystems,
since they hold memory for that vnodes. For instance, perfused grows
and cr
On Tue, Jun 05, 2012 at 10:10:30PM +0200, Hans Petter Selasky wrote:
> If two SMP enabled stacks have each their lock, and they are
> calling each other, that means any callbacks must go unlocked,
> because else you can get a LOR (locking order reversal). Agree?
No. If this is an issue, the loc