Hi all,
I still use the following kickstart partition scheme for C7 installations (via virt-install): Briefly, fixed size for /root and /boot, and the rest is filled up for /srv.
The same kickstart (despite that c7 uses vda, f29 uses sda) doesn't work with Fedora29 (EL8). I get a "device is too small for new format" error. Any hints?
part /RESCUE --fstype="ext4" --ondisk=sda --size=1280 --label=RESCUE part /boot --fstype="ext4" --ondisk=sda --size=512 --label=BOOT part pv.0104 --fstype="lvmpv" --ondisk=sda --grow volgroup ee --pesize=4096 pv.1974 logvol / --fstype="ext4" --size=3072 --encrypted --label="ROOT" --name=00 --vgname=ee logvol /srv --fstype="ext4" --percent=100 --grow --encrypted --label="SRV" --name=01 --vgname=ee
Thanks, LF
Leon Fauster via CentOS wrote:
Hi all,
I still use the following kickstart partition scheme for C7 installations (via virt-install): Briefly, fixed size for /root and /boot, and the rest is filled up for /srv.
The same kickstart (despite that c7 uses vda, f29 uses sda) doesn't work with Fedora29 (EL8). I get a "device is too small for new format" error. Any hints?
part /RESCUE --fstype="ext4" --ondisk=sda --size=1280 --label=RESCUE part /boot --fstype="ext4" --ondisk=sda --size=512 --label=BOOT part pv.0104 --fstype="lvmpv" --ondisk=sda --grow volgroup ee --pesize=4096 pv.1974 logvol / --fstype="ext4" --size=3072 --encrypted --label="ROOT" --name=00 --vgname=ee logvol /srv --fstype="ext4" --percent=100 --grow --encrypted --label="SRV" --name=01 --vgname=ee
If I'm reading that correctly - haven't worked on a kickstart in years - I'd start by increasing root to 1024 (1M) for /boot.
mark
On Wed, 8 May 2019 at 13:24, mark m.roth@5-cent.us wrote:
Leon Fauster via CentOS wrote:
Hi all,
I still use the following kickstart partition scheme for C7 installations (via virt-install): Briefly, fixed size for /root and /boot, and the rest is filled up for /srv.
The same kickstart (despite that c7 uses vda, f29 uses sda) doesn't work with Fedora29 (EL8). I get a "device is too small for new format" error. Any hints?
part /RESCUE --fstype="ext4" --ondisk=sda --size=1280 --label=RESCUE part /boot --fstype="ext4" --ondisk=sda --size=512 --label=BOOT part pv.0104 --fstype="lvmpv" --ondisk=sda --grow volgroup ee
--pesize=4096
pv.1974 logvol / --fstype="ext4" --size=3072 --encrypted --label="ROOT" --name=00 --vgname=ee logvol /srv --fstype="ext4" --percent=100 --grow --encrypted
--label="SRV"
--name=01 --vgname=ee
If I'm reading that correctly - haven't worked on a kickstart in years - I'd start by increasing root to 1024 (1M) for /boot.
Yep. Minimum for that is going to be about the same as your RESCUE. The other would just be to confirm that the sda has space and nothing still on it which it was trying to work around.
mark
CentOS mailing list CentOS@centos.org https://lists.centos.org/mailman/listinfo/centos
On 2019-05-08 12:28, Stephen John Smoogen wrote:
On Wed, 8 May 2019 at 13:24, mark m.roth@5-cent.us wrote:
Leon Fauster via CentOS wrote:
Hi all,
I still use the following kickstart partition scheme for C7 installations (via virt-install): Briefly, fixed size for /root and /boot, and the rest is filled up for /srv.
The same kickstart (despite that c7 uses vda, f29 uses sda) doesn't work with Fedora29 (EL8). I get a "device is too small for new format" error. Any hints?
part /RESCUE --fstype="ext4" --ondisk=sda --size=1280 --label=RESCUE part /boot --fstype="ext4" --ondisk=sda --size=512 --label=BOOT part pv.0104 --fstype="lvmpv" --ondisk=sda --grow volgroup ee
--pesize=4096
pv.1974 logvol / --fstype="ext4" --size=3072 --encrypted --label="ROOT" --name=00 --vgname=ee logvol /srv --fstype="ext4" --percent=100 --grow --encrypted
--label="SRV"
--name=01 --vgname=ee
If I'm reading that correctly - haven't worked on a kickstart in years - I'd start by increasing root to 1024 (1M) for /boot.
Yep. Minimum for that is going to be about the same as your RESCUE. The other would just be to confirm that the sda has space and nothing still on it which it was trying to work around.
In CentOS releases 5 --> 6 --> 7
the demand for /boot size doubles if not triples with each release... Otherwise one day you may fill it up before number of kernels will be such that kernel update will remove oldest kernel. And _we_ called Windows 2000 "bloated pig" when it was released... Sigh.
Valeri
mark
CentOS mailing list CentOS@centos.org https://lists.centos.org/mailman/listinfo/centos
On Wed, 8 May 2019 at 13:48, Valeri Galtsev galtsev@kicp.uchicago.edu wrote:
On 2019-05-08 12:28, Stephen John Smoogen wrote:
On Wed, 8 May 2019 at 13:24, mark m.roth@5-cent.us wrote:
Yep. Minimum for that is going to be about the same as your RESCUE. The other would just be to confirm that the sda has space and nothing still
on
it which it was trying to work around.
In CentOS releases 5 --> 6 --> 7
the demand for /boot size doubles if not triples with each release... Otherwise one day you may fill it up before number of kernels will be such that kernel update will remove oldest kernel. And _we_ called Windows 2000 "bloated pig" when it was released... Sigh.
Software seems to grow to the maximum space it can occupy. I think in 1989 we were complaining about BSD not being able to fit on our VAX 750's boot drive anymore and we needed to put in a 40MB drive system instead. I expect by the 2040's we will be looking at petabyte drives and wondering how we can fit anything on it.
Valeri
mark
CentOS mailing list CentOS@centos.org https://lists.centos.org/mailman/listinfo/centos
-- ++++++++++++++++++++++++++++++++++++++++ Valeri Galtsev Sr System Administrator Department of Astronomy and Astrophysics Kavli Institute for Cosmological Physics University of Chicago Phone: 773-702-4247 ++++++++++++++++++++++++++++++++++++++++ _______________________________________________ CentOS mailing list CentOS@centos.org https://lists.centos.org/mailman/listinfo/centos
On 2019-05-08 15:05, Stephen John Smoogen wrote:
On Wed, 8 May 2019 at 13:48, Valeri Galtsev galtsev@kicp.uchicago.edu wrote:
On 2019-05-08 12:28, Stephen John Smoogen wrote:
On Wed, 8 May 2019 at 13:24, mark m.roth@5-cent.us wrote:
Yep. Minimum for that is going to be about the same as your RESCUE. The other would just be to confirm that the sda has space and nothing still
on
it which it was trying to work around.
In CentOS releases 5 --> 6 --> 7
the demand for /boot size doubles if not triples with each release... Otherwise one day you may fill it up before number of kernels will be such that kernel update will remove oldest kernel. And _we_ called Windows 2000 "bloated pig" when it was released... Sigh.
Software seems to grow to the maximum space it can occupy.
Of course, this is true. I do remember big upgrade: 40 MB hard drive replacing 20 MB one. Still, during the same decade an a half covering CentOS releases I mentioned, on my FreeBSD boxes /boot grew up less than twice, whereas on my CentOS boxes it grew up at least 5 times. I safely run CentOS 5 through its whole lifespan with /boot as small as 200MB. On CentOS 7 I make /boot 3 GB (maybe slight overkill, but boxes with 1 GB went into /boot size issue after several kernel updates). But what one can do: Linux kernel has a lot of stuff that, hmm..., one can probably live without, and what comes with Linux distributions covers widest variety of hardware it will run on ;-)
Valeri PS Yes, I run and programmed for machines with 4 - 16 Kb of RAM, and now administer machine with almost 1 TB of RAM. Now you can pretty much pinpoint my age ;-)
I think in 1989 we were complaining about BSD not being able to fit on our VAX 750's boot drive anymore and we needed to put in a 40MB drive system instead. I expect by the 2040's we will be looking at petabyte drives and wondering how we can fit anything on it.
Valeri
mark
CentOS mailing list CentOS@centos.org https://lists.centos.org/mailman/listinfo/centos
-- ++++++++++++++++++++++++++++++++++++++++ Valeri Galtsev Sr System Administrator Department of Astronomy and Astrophysics Kavli Institute for Cosmological Physics University of Chicago Phone: 773-702-4247 ++++++++++++++++++++++++++++++++++++++++ _______________________________________________ CentOS mailing list CentOS@centos.org https://lists.centos.org/mailman/listinfo/centos
On Wed, 8 May 2019 at 17:48, Valeri Galtsev galtsev@kicp.uchicago.edu wrote:
On 2019-05-08 15:05, Stephen John Smoogen wrote:
On Wed, 8 May 2019 at 13:48, Valeri Galtsev galtsev@kicp.uchicago.edu wrote:
On 2019-05-08 12:28, Stephen John Smoogen wrote:
On Wed, 8 May 2019 at 13:24, mark m.roth@5-cent.us wrote:
Yep. Minimum for that is going to be about the same as your RESCUE. The other would just be to confirm that the sda has space and nothing still
on
it which it was trying to work around.
In CentOS releases 5 --> 6 --> 7
the demand for /boot size doubles if not triples with each release... Otherwise one day you may fill it up before number of kernels will be such that kernel update will remove oldest kernel. And _we_ called Windows 2000 "bloated pig" when it was released... Sigh.
Software seems to grow to the maximum space it can occupy.
Of course, this is true. I do remember big upgrade: 40 MB hard drive replacing 20 MB one. Still, during the same decade an a half covering CentOS releases I mentioned, on my FreeBSD boxes /boot grew up less than twice, whereas on my CentOS boxes it grew up at least 5 times. I safely run CentOS 5 through its whole lifespan with /boot as small as 200MB. On CentOS 7 I make /boot 3 GB (maybe slight overkill, but boxes with 1 GB went into /boot size issue after several kernel updates). But what one can do: Linux kernel has a lot of stuff that, hmm..., one can probably live without, and what comes with Linux distributions covers widest variety of hardware it will run on ;-)
I never know how much of the growth differences is part and parcel of why Linux is seen more often than FreeBSD.. or an unwanted side effect that gets used as an excuse for why it is used more often. I do know that the Linux distributions which focus on smallness usually find themselves only in niche roles because they don't have all the bells and whistles someone expected from a larger version. They then find themselves fighting off an existing base of *BSD systems which have been in that small/tidy base for decades.
Back to technical terms.. the vast amount of space being used on a /boot partition is the initial ram disks which contain mostly user space tools. In a 'weird' way you could say you are seeing a 'micro-kernel' approach where various things which would have been in the kernel decades ago are pushed off to user-space.. however because the system needs those to be up very early (aka why they were in the kernel long ago) you push it into the initial ram disk. So the old school Unix people can say that Linux distributions are doing Unix kernel space wrong, and the old school micro-kernel people can say we are doing Micro-kernels wrong.
Or they can say we are doing it right.. because it got moved out of the kernel to make sure the kernel only does specific things versus a MULTIX way.. and the microkernel people get to see what user space utilities .. Either way lots of pub-geek fight material.
Valeri PS Yes, I run and programmed for machines with 4 - 16 Kb of RAM, and now administer machine with almost 1 TB of RAM. Now you can pretty much pinpoint my age ;-)
Oh no no. I am not falling for that trap. Never ask another person on the internet their age.. you will either find out you are much much younger than them and so should have known better or much much older and should have known better.
I think in 1989 we were complaining about BSD not being able to fit on our VAX 750's boot drive anymore and we needed to put in a 40MB drive system instead. I
expect
by the 2040's we will be looking at petabyte drives and wondering how we can fit anything on it.
Valeri
mark
CentOS mailing list CentOS@centos.org https://lists.centos.org/mailman/listinfo/centos
-- ++++++++++++++++++++++++++++++++++++++++ Valeri Galtsev Sr System Administrator Department of Astronomy and Astrophysics Kavli Institute for Cosmological Physics University of Chicago Phone: 773-702-4247 ++++++++++++++++++++++++++++++++++++++++ _______________________________________________ CentOS mailing list CentOS@centos.org https://lists.centos.org/mailman/listinfo/centos
-- ++++++++++++++++++++++++++++++++++++++++ Valeri Galtsev Sr System Administrator Department of Astronomy and Astrophysics Kavli Institute for Cosmological Physics University of Chicago Phone: 773-702-4247 ++++++++++++++++++++++++++++++++++++++++ _______________________________________________ CentOS mailing list CentOS@centos.org https://lists.centos.org/mailman/listinfo/centos
Once upon a time, Stephen John Smoogen smooge@gmail.com said:
Software seems to grow to the maximum space it can occupy. I think in 1989 we were complaining about BSD not being able to fit on our VAX 750's boot drive anymore and we needed to put in a 40MB drive system instead. I expect by the 2040's we will be looking at petabyte drives and wondering how we can fit anything on it.
I first installed Linux on a 386SX (16MHz IIRC) system with a pair of 20MB hard drives and 2MB RAM... it took a little while. :) I think we ended up just using one drive for the root filesystem and the other for swap! This was probably 1992 or 1993, can't remember.
Today I installed Linux on a system wtih 48 CPU cores...
Am 08.05.2019 um 19:20 schrieb mark m.roth@5-cent.us:
Leon Fauster via CentOS wrote:
Hi all,
I still use the following kickstart partition scheme for C7 installations (via virt-install): Briefly, fixed size for /root and /boot, and the rest is filled up for /srv.
The same kickstart (despite that c7 uses vda, f29 uses sda) doesn't work with Fedora29 (EL8). I get a "device is too small for new format" error. Any hints?
part /RESCUE --fstype="ext4" --ondisk=sda --size=1280 --label=RESCUE part /boot --fstype="ext4" --ondisk=sda --size=512 --label=BOOT part pv.0104 --fstype="lvmpv" --ondisk=sda --grow volgroup ee --pesize=4096 pv.0104 logvol / --fstype="ext4" --size=3072 --encrypted --label="ROOT" --name=00 --vgname=ee logvol /srv --fstype="ext4" --percent=100 --grow --encrypted --label="SRV" --name=01 --vgname=ee
If I'm reading that correctly - haven't worked on a kickstart in years - I'd start by increasing root to 1024 (1M) for /boot.
That sounds reasonable but it doesn't help (tested up to 3GB for boot). The point is, that with fixed sizes everything works (even with 3GB root, 512MB boot, 768 RAM). The anaconda text-installation process passes without any problems. Only when "--grow" option comes in the problem appears. And this only for Fedora29, a C7 installation works fine (dynamically expands to the boundaries of the disk).
Looking into the docs doesn't show any modifications: https://pykickstart.readthedocs.io/en/latest/kickstart-docs.html#chapter-3-k...
It seems that with every new major release the headaches starts again :-(.
<retest>
While writing I did a new installation; the grow option for the "part" command seems to work. The VG expands to the disk boundaries. So, the problem gets more located around the "logvol" command ...
-- LF
Once upon a time, Leon Fauster via CentOS centos@centos.org said:
And this only for Fedora29, a C7 installation works fine (dynamically expands to the boundaries of the disk).
You are asking about Fedora - you'd probably get better results on the Fedora mailing lists, forums, etc.
Am 08.05.2019 um 21:12 schrieb Chris Adams linux@cmadams.net:
Once upon a time, Leon Fauster via CentOS centos@centos.org said:
And this only for Fedora29, a C7 installation works fine (dynamically expands to the boundaries of the disk).
You are asking about Fedora - you'd probably get better results on the Fedora mailing lists, forums, etc.
Well, F29 is very close to EL8 but you are right. Anyway, for the sake of precision. I made the same installation with RHEL8 (anaconda 29.19 vs 29.24 on F29). The results are same!
Following "logvol" stanza "--percent=100 --grow" works for EL7 but not for EL8.
While doing this again I found a solution; "--size=SIZE --grow" as options with an arbitrary SIZE will grow the LV to the size of the free VG size.
-- LF
Am 08.05.2019 um 20:56 schrieb Markus Falb markus.falb@fasel.at:
On 08.05.19 19:16, Leon Fauster via CentOS wrote: ...
I get a "device is too small for new format" error. Any hints?
...> part pv.0104 --fstype="lvmpv" --ondisk=sda --grow
pv.0104
volgroup ee --pesize=4096 pv.1974
pv.1974
mismatch!
Yep, I saw it after sending the mail. Its just a cut and paste and cleanup mistake. The ks file has the matching ids correct. Thanks.
-- LF
-
Chris Adams -
Leon Fauster -
mark -
Markus Falb -
Stephen John Smoogen -
Valeri Galtsev