>>> On 12/19/2014 at 02:54 PM, Mark Pace wrote:
> So during install I don't see a way to add 3 disks to btrfs without first
> creating an LVM volume group. I see no way to run this command, btrfs
> device add dev1 dev2 /path/to/filesystem during install.
I haven't had the chance to play around
I'm going back and putting / in a partition of it's own.
And will then try to add the other 3 disks to a btrfs.
So during install I don't see a way to add 3 disks to btrfs without first
creating an LVM volume group. I see no way to run this command, btrfs
device add dev1 dev2 /path/to/filesystem
>>> On 12/19/2014 at 09:46 AM, Mark Pace wrote:
> sles003:~> df -Ht
> df: option requires an argument -- 't'
> Try 'df --help' for more information.
> sles003:~> df -H
> Filesystem Size Used Avail Use% Mounted on
> /dev/mapper/system-root 5.7G 2.5G 3.0G 46% /
> devtmpfs
On 18 December 2014 at 20:26, Mark Post wrote:
>
> >>> On 12/18/2014 at 09:32 AM, "Levy, Alan" wrote:
> > Does it make sense in just setting up one mod 29 (32000+ cylinders, app
> 21G)
> > and let the btrfs use the whole thing instead of breaking it up into
> smaller
> > pieces (var, opt, home, e
sles003:~> df -Ht
df: option requires an argument -- 't'
Try 'df --help' for more information.
sles003:~> df -H
Filesystem Size Used Avail Use% Mounted on
/dev/mapper/system-root 5.7G 2.5G 3.0G 46% /
devtmpfs 460M 8.2k 460M 1% /dev
tmpfs467
were lucky in the regard.
Harley
-Original Message-
From: Linux on 390 Port [mailto:LINUX-390@VM.MARIST.EDU] On Behalf Of Offer
Baruch
Sent: Thursday, December 18, 2014 11:06 PM
To: LINUX-390@VM.MARIST.EDU
Subject: Re: SLES12 / partitioning
Hi
There is a good reason to split some director
Hi
There is a good reason to split some directories.
/home if set on / fs then any user can fill up your / fs and fail the
server.
/var might fill up very quickly in different logging bursts or application
errors or loops and fail the entire server. No matter how much free space
its got
Same for /
>>> On 12/18/2014 at 09:32 AM, "Levy, Alan" wrote:
> Does it make sense in just setting up one mod 29 (32000+ cylinders, app 21G)
> and let the btrfs use the whole thing instead of breaking it up into smaller
> pieces (var, opt, home, etc) ?
Without anything to back this up, my concern is that
>>> On 12/18/2014 at 09:54 AM, Mark Pace wrote:
> Playing around with the Installer and the DASD layout I found the setting
> to use a LVM based btrfs. This created a partition with ext2 for the
> /boot/zipl and then created a partition for LVM and btrfs for the root. I
> then added the rest of
On 12/17/2014 10:16 AM, Michael MacIsaac wrote:
>> > You might also consider TMPFS for /tmp. It's much faster if you don't
>> > need huge /tmp capacity or persistence.
> At one point, we started to recommend this in "The Virtualization
> Cookbook". The reply was to avoid it because tmpfs costs you
PM
> To: LINUX-390@VM.MARIST.EDU
> Subject: Re: SLES12 / partitioning
>
> >>> On 12/17/2014 at 09:41 AM, Mark Pace wrote:
> > With btrfs now the default file system on sles12, is there a new
> > recommended way to layout DASD devices?
> > I would think if you had rea
Sent: Wednesday, December 17, 2014 6:10 PM
To: LINUX-390@VM.MARIST.EDU
Subject: Re: SLES12 / partitioning
>>> On 12/17/2014 at 09:41 AM, Mark Pace wrote:
> With btrfs now the default file system on sles12, is there a new
> recommended way to layout DASD devices?
> I would think
>>> On 12/17/2014 at 10:16 AM, Michael MacIsaac wrote:
> Rick,
>
>> You might also consider TMPFS for /tmp. It's much faster if you don't
>> need huge /tmp capacity or persistence.
> At one point, we started to recommend this in "The Virtualization
> Cookbook". The reply was to avoid it because
>>> On 12/17/2014 at 10:09 AM, Rick Troth wrote:
> Looks fine, Mark.
> Isolating root (or at least /boot) is a great idea.
In the case of btrfs, having /boot/zipl a separate file system which is _not_
btrfs is required on System z. The zipl command and IPL code get unhappy if
that is not the
>>> On 12/17/2014 at 09:41 AM, Mark Pace wrote:
> With btrfs now the default file system on sles12, is there a new
> recommended way to layout DASD devices?
> I would think if you had really big 3390 devices that btrfs makes sense.
> But still working with mod-9s it doesn't seem like the right fi
Rick,
> You might also consider TMPFS for /tmp. It's much faster if you don't
> need huge /tmp capacity or persistence.
At one point, we started to recommend this in "The Virtualization
Cookbook". The reply was to avoid it because tmpfs costs you memory which
is much more valuable on z than disk.
On 12/17/2014 09:41 AM, Mark Pace wrote:
> With btrfs now the default file system on sles12, is there a new
> recommended way to layout DASD devices?
As long as you're using normal LVM, there is no need to change your layout.
> I would think if you had really big 3390 devices that btrfs makes se
Mark,
> With btrfs now the default file system on sles12, is there a new
> recommended way to layout DASD devices?
Good question.
I have also been thinking of the Copy On Write (COW) feature. So a
follow-up question might be: "How do I utilize the cloning and COW features
of btrfs?" .
With cont
With btrfs now the default file system on sles12, is there a new
recommended way to layout DASD devices?
I would think if you had really big 3390 devices that btrfs makes sense.
But still working with mod-9s it doesn't seem like the right fit.
I'm still putting / on a minidisk and then creating a
>>> On 11/19/2014 at 07:43 AM, "Levy, Alan" wrote:
> During the install, I removed /var, /opt/ ,tmp and /home from the btrfs
> (dasda). I then created these partitions on their own device (dasdb thru
> dasde). When I log in, via maintenance mode due to the failed dependency
> errors, everythin
@VM.MARIST.EDU
Subject: Re: SLES12 / partitioning
>>> On 11/12/2014 at 08:46 AM, "Levy, Alan" wrote:
> Just starting to test with sles12. My colleagues and I are trying to
> find a good way of setting up the mount points.
>
> In the past, we have used the fo
>>> On 11/12/2014 at 08:46 AM, "Levy, Alan" wrote:
> Just starting to test with sles12. My colleagues and I are trying to find a
> good way of setting up the mount points.
>
> In the past, we have used the following methodology (on a mod-9):
>
> 191 - profile exec
> 292 - 1500 cyl/1G
Just starting to test with sles12. My colleagues and I are trying to find a
good way of setting up the mount points.
In the past, we have used the following methodology (on a mod-9):
191 - profile exec
292 - 1500 cyl/1G /
293 - 730 cyl/.5G swap
294 - 4785 cyl/3.3G
23 matches
Mail list logo
