Possibly, but the discussion of specific features has barely been broached.
On Wednesday, March 12, 2014 5:42:40 AM UTC-7, David Cantrell wrote:
>
> On Mon, Mar 10, 2014 at 05:01:58PM -0700, Jorgen Lundman wrote:
>
> > Once the installer is pushed out for the current "stable" version, the
> next
On Mon, Mar 10, 2014 at 05:01:58PM -0700, Jorgen Lundman wrote:
> Once the installer is pushed out for the current "stable" version, the next
> focus is on OSX integration
Will this include GUI management of pools/volumes/etc?
--
David Cantrell | Minister for Arbitrary Justice
Safety tip:
We have no at all looked at performance, since focus was to get it working,
then to get it stable. On the hangs, we hope to have already fixed them. I
just tagged ZFS so the version should say
ZFS: Loaded module v0.6.2-rc1, ZFS pool version 5000, ZFS filesystem
version 5
As long as you don't c
Hey Jorgen, this compilation was a pull that is still considered current
(alpha-276-ga87b684). What can I do to better track these issues down
presently for you? I'm about to take a jump from ZFS due to these issues,
so I'd be happy to help if we can get to the bottom of them before I give
up
Once the installer is pushed out for the current "stable" version, the next
focus is on OSX integration, and performance improvements. Do we have a
feel for how the latest snapshots performs compared to ZEVO? Are we talking
5-10-20%, or is it so high we suspect there is a problem?
Having said
yllluminate,
Yes its a given that we are slower that Zevo at this point.
I am disturbed to hear of your hangs. Perhaps you could be a bit more
specific and let us know what build of zfs-osx you are running, on what
hardware, and importantly whether the machine can be sshed into when it
enters
Yeah, I certainly wish it was just performance. A hang such as this is
well beyond a performance issue at the moment. I'm probably going to just
go back to using ZFS on FBSD to keep my backups and use software RAID on my
workstations until maczfs matures or Greenbytes get's its proverbial head
I also switched from Zevo this weekend. I didn't want to be stuck on 10.8
forever and I noticed some performance difference, that's for sure.
However, I'm grateful for all the work the guys are doing so I can use it
:) Hopefully performance will be addressed too though.
On Sun, Mar 9, 2014 at 3:
Well I'm getting hangs that are at least an order of magnitude worse than
ZEVO now. At one point prior to the memory issues being worked around I
had a lot of hangs where I would have to reboot the system from a remote
ssh session or just hold the power in and hard reboot it. I am finding
aft
Haha. I'm not claiming anything other than results of current testing. However
for me, importing old pools and updating has not provided the performance of
starting fresh. All that said, lot of things could be different in individual
pool creation. Those coming from ZEVO may have significantly d
Is my understanding correct that you can't "fix" a pool after it's been
created, but you actually have to create a new pool with the optimal
settings of the day, and copy data around to get performance improvements?
That doesn't sound right.
On Sun, Mar 9, 2014 at 6:43 AM, Jason Belec wrote:
>
Haha, it was not derogatory just business term for guys coming out of Apple,
starting something with the tech and getting acquired for whatever reason. Been
there myself. However it does not change how end users feel. ;)
As for your options, the change seems to be for pool vs zfs directory, but
Please do not say things like that about Don. I know him personally and
that is not what happened. I'm not going to get into details here as it is
not appropriate, but it is frankly not what happened.
I'll continue on my path then and work on getting everything off this pool
and recreate it.
Yes to everything you noticed.
That said, hoping for anything Don did, is utterly pointless. He walked away
with cash in his pocket.
Currently pools are being built with commands like this
zpool create -f -O compression=lz4 -O casesensitivity=insensitive -O
normalization=formD -O atime=off
Also curious, what has been the best pool and zfs creation args you've seen
so far with best results here? After I get data migrated off of this pool
and onto a temp volume I'll recreate the pool and run a test again here in
a few days to see if I get some different results. Nonetheless, I
ce
Here's a history( `zpool history`):
https://gist.github.com/915f6d99b219a80d7cdb
--
---
You received this message because you are subscribed to the Google Groups
"zfs-macos" group.
To unsubscribe from this group and stop receiving emails from it, send an email
to zfs-macos+unsubscr...@googleg
Whats the pool history say about how this was created?
I understand your importing, which may contribute to issues. Seen some in pools
that are old that I've imported before copying contents to new pools with the
suggested parameters.
--
Jason Belec
Sent from my iPad
> On Mar 8, 2014, at 8:5
Oh, and just to give a quick idea of how dramatic those peaks are compared
to the last 24 hours without zfs enabled, here is a graph:
http://cl.ly/image/3f2e2t1w2s38
--
---
You received this message because you are subscribed to the Google Groups
"zfs-macos" group.
To unsubscribe from this gr
In this particular case the pool is an imported ZEVO pool. I don't yet
have the luxury of creating a new pool until I move this data around
I did perform an upgrade on this pool and here are the settings:
zpool: https://gist.github.com/25315011067796c62b43
zfs: https://gist.github.com/03d7016422
Can you give us the info that you used to build the test pools?
--
Jason Belec
Sent from my iPad
> On Mar 8, 2014, at 8:28 PM, ylluminate wrote:
>
> Hi folks, thought it was time to try maczfs again. I am shifting over from
> ZEVO due to GreenByte's fubar of the project. Seems like the shi
Hi folks, thought it was time to try maczfs again. I am shifting over from
ZEVO due to GreenByte's fubar of the project. Seems like the shift over to
OpenZFS is worthwhile and I'm satisfied to see some positive results.
One thing that has bothered me so far is the marked differential in spee
21 matches
Mail list logo
