I think I installed Opensuse 11.x a few years ago with the standard suggested partition, more or less.
After that I have re installed but kept the partitioning and data.
Now I have come to a situation where I can’t install the latest kernel update because /boot is too small.
My problem is that the rest is used for LVM.
Is it possible to solve this without reinstalling and repartitioning everything?
This is how it looks from fdisk:
Disk /dev/sda: 1000.2 GB, 1000204886016 bytes
255 heads, 63 sectors/track, 121601 cylinders, total 1953525168 sectors
Units = sectors of 1 * 512 = 512 bytes
Sector size (logical/physical): 512 bytes / 512 bytes
I/O size (minimum/optimal): 512 bytes / 512 bytes
Disk identifier: 0x0000e7cf
Device Boot Start End Blocks Id System
/dev/sda1 * 63 144584 72261 83 Linux
/dev/sda2 144585 31583789 15719602+ 8e Linux LVM
/dev/sda3 31583790 1953520064 960968137+ 8e Linux LVM
It would be nice to re allocate sda1 and sda2 to /boot ( or give at least 1 GB of sda2 to sda1).
to do that I need to make sure lvm only uses sda3, remove sda2 from the lvm and resize.
How can this be done?
A backup plan is to do a full backup to my NAS and reinstall/repartition but that takes time…
I do have unallocated space, my problem is that I have 4 volumes that span over the 2 partitions (physical volumes).
How do I move the volumes ?
this is my lvdisplay output :
— Logical volume —
LV Path /dev/system/home
LV Name home
VG Name system
LV UUID Ru1glv-d0Ep-3LdY-uB7R-pTR8-aWwg-811EDv
LV Write Access read/write
LV Creation host, time ,
LV Status available
open 1
LV Size 200.00 GiB
Current LE 51200
Segments 2
Allocation inherit
Read ahead sectors auto
currently set to 256
Block device 253:0
— Logical volume —
LV Path /dev/system/root
LV Name root
VG Name system
LV UUID Cz4N0F-VXSI-Q0Of-xtYg-3YfW-gdnC-n2eIR1
LV Write Access read/write
LV Creation host, time ,
LV Status available
open 1
LV Size 20.00 GiB
Current LE 5120
Segments 2
Allocation inherit
Read ahead sectors auto
currently set to 256
Block device 253:1
— Logical volume —
LV Path /dev/system/swap
LV Name swap
VG Name system
LV UUID JVBKZ9-JYKm-Ej6Y-jodK-TEXC-9Kb0-BcA0kw
LV Write Access read/write
LV Creation host, time ,
LV Status available
open 2
LV Size 2.00 GiB
Current LE 512
Segments 1
Allocation inherit
Read ahead sectors auto
currently set to 256
Block device 253:2
— Logical volume —
LV Path /dev/system/data
LV Name data
VG Name system
LV UUID RrNynG-wWRj-AgEw-pNs6-nQnU-LxMc-0cLHxp
LV Write Access read/write
LV Creation host, time ,
LV Status available
open 1
LV Size 400.00 GiB
Current LE 102400
Segments 1
Allocation inherit
Read ahead sectors auto
On 05/20/2013 11:06 AM, joakim ganse wrote:
> Now I have come to a situation where I can’t install the latest kernel
> update because /boot is too small.
> My problem is that the rest is used for LVM.
i’d ‘argue’ the problem is that you now have several kernels and some
systemd ‘stuff’ in a partition originally sized for one kernel and
systemv…i guess you could delete all kernels but the one now
running good before adding the new one…and maybe skinny down some
systemd stuff…
caution: i’ve not mentioned you might need to do some surgery on the
systemd/grub2 stuff (which i don’t understand and won’t now try
to)…so, make sure you know how to remove unneeded kernels and ???
(some systemd thing that grows).
read my caveat and wait for a systemd/grub2 wizard before hacking.
Please, please, please
Post all computer text between CODE tags. You getthe tags by clicking on the # button in the toolbar of the post editor. And then copy/paste from your terminal window the prompt, the command, the output and the next prompe in between those Code tags.
Also, before declaring your /boot patition “full”, please show what is in there. As said, there could be leftover kernels and deleting those might be a much better solution then reorganising the whole setup.
hcvv wrote:
> Please, please, please
> Post all computer text between CODE tags. You getthe tags by clicking
> on the # button in the toolbar of the post editor. And then copy/paste
> from your terminal window the prompt, the command, the output and the
> next prompe in between those Code tags.
That’s generally good advice, but there’s no problem reading Joakim’s
post in this case. More problematic is that it doesn’t contain enough
information.
> Also, before declaring your /boot patition “full”, please show what is
> in there. As said, there could be leftover kernels and deleting those
> might be a much better solution then reorganising the whole setup.
That is correct (although I feel a slight headache :)), but as it is difficult for new posters to find out, I try to explain this for the benefit of future posting.
On 2013-05-20 11:59, Dave Howorth wrote:
> hcvv wrote:
>> Please, please, please
>> Post all computer text between CODE tags. You getthe tags by clicking
>> on the # button in the toolbar of the post editor. And then copy/paste
>> from your terminal window the prompt, the command, the output and the
>> next prompe in between those Code tags.
>
> That’s generally good advice, but there’s no problem reading Joakim’s
> post in this case. More problematic is that it doesn’t contain enough
> information.
It is not a problem for you and me, via nntp, but it is for him via http
because tabulation is destroyed. It could be even worse, with smileys,
but not this once.
–
Cheers / Saludos,
Carlos E. R.
(from 12.1 x86_64 “Asparagus” at Telcontar)
lets see.’
Did some investigation meanwhile and have come “halfway”.
Used pvdisplay /dev/sda2 to see that the physical volume was used ( knew that already, but anyway)
Then used pvmove /dev/sda2 /dev/sda3 to move the stuff over followed by vgreduce system /dev/sda2 to remove sda2 from the system volume.
So, now I have a non used sda2.
This is what I have right now:
vgscan:
Reading all physical volumes. This may take a while...
Found volume group "system" using metadata type lvm2
pvscan
PV /dev/sda3 VG system lvm2 [916.45 GiB / 294.45 GiB free]
PV /dev/sda2 lvm2 [14.99 GiB]
Total: 2 [931.44 GiB] / in use: 1 [916.45 GiB] / in no VG: 1 [14.99 GiB]
lvscan:
PV /dev/sda3 VG system lvm2 [916.45 GiB / 294.45 GiB free]
PV /dev/sda2 lvm2 [14.99 GiB]
Total: 2 [931.44 GiB] / in use: 1 [916.45 GiB] / in no VG: 1 [14.99 GiB]
boven:~ # du -sh /boot
38M /boot
boven:~ # l /boot
total 30232
drwxr-xr-x 4 root root 4096 Mar 15 17:58 ./
drwxr-xr-x 24 root root 4096 May 20 09:30 ../
-rw-r--r-- 1 root root 2490431 Feb 27 15:58 System.map-3.4.33-2.24-desktop
-rw------- 1 root root 512 Dec 11 12:15 backup_mbr
lrwxrwxrwx 1 root root 1 Dec 11 12:04 boot -> ./
-rw-r--r-- 1 root root 1236 Jul 15 2012 boot.readme
-rw-r--r-- 1 root root 131335 Feb 27 14:59 config-3.4.33-2.24-desktop
drwxr-xr-x 2 root root 4096 Dec 11 12:04 grub/
drwxr-xr-x 6 root root 4096 Mar 15 17:59 grub2/
lrwxrwxrwx 1 root root 26 Mar 15 17:58 initrd -> initrd-3.4.33-2.24-desktop
-rw-r--r-- 1 root root 16836836 Mar 15 17:58 initrd-3.4.33-2.24-desktop
-rw-r--r-- 1 root root 581632 Dec 11 12:48 message
-rw-r--r-- 1 root root 222481 Feb 27 16:09 symvers-3.4.33-2.24-desktop.gz
-rw-r--r-- 1 root root 516 Feb 27 16:09 sysctl.conf-3.4.33-2.24-desktop
-rw-r--r-- 1 root root 5725835 Feb 27 16:09 vmlinux-3.4.33-2.24-desktop.gz
lrwxrwxrwx 1 root root 27 Mar 15 17:57 vmlinuz -> vmlinuz-3.4.33-2.24-desktop
-rw-r--r-- 1 root root 4924144 Feb 27 15:58 vmlinuz-3.4.33-2.24-desktop
boven:~ #
All about the same size ecept for the initrd That si about twice as large as mine. And 16 MB out of 69 does count. And you can never accomodate two kernels there. Thus I agree that enlarging it mightt be best.
PS, do you see above why I asked you to copy/past including the command? Then you do not have to explain that you did an “ll on /boot” because it is already there. And it does not only save you that explanation, we can then see exactly what you did and not hope that your explanation is correct.
Yes, I see what you mean with the code wrap, makes it easier for the eye.
What I have done so far is removed sdb2 from the volume set so right now nothing is using it.
It is still of type 8e, lvm that is.
Is is safe to just delete it and then extend /boot wit all or parts of it ?
I don’t think anything really will be broken, just wondering if the system will handle the automatic change when (i assume) sda3 suddenly becomes sda2 for volume group system.
I was generous when I installed SuSE 10.2. I allowed 100M for “/boot”. And now people are asking why I made it so small
In my case, I deleted plymouth. That reduces the size of the “initrd”. With two kernels, I’m using around 64M of space on “/boot”. The other option is to edit “/etc/zypper/zypp.conf” and comment out the option for multiversion kernels. But note that I have not tested whether that will work.
For my next install on that box, I will probably delete WinXP (which is near end-of-life), and give some of its partition to “/boot”.
On 2013-05-20 15:46, nrickert wrote:
>
> joakim_ganse;2558503 Wrote:
>> I think I installed Opensuse 11.x a few years ago with the standard
>> suggested partition, more or less.
>> After that I have re installed but kept the partitioning and data.
>
> I was generous when I installed SuSE 10.2. I allowed 100M for “/boot”.
> And now people are asking why I made it so small
I have the same problem - mine is 190 MiB. I made it BIG!
(I will no longer joke about 540KiB being enough memory for everybody…)
> In my case, I deleted plymouth.
I’m testing exactly that in a virtual machine, to figure out how to do
it and use encrypted partitions…
–
Cheers / Saludos,
Carlos E. R.
(from 12.1 x86_64 “Asparagus” at Telcontar)
Oops. That should have been SuSE 10.1, which first installed back in 2006 on hardware that has been sent to the graveyard for computers. I reinstalled on a replacement box in 2007, and that’s when I set “/boot” to 100M. I seem to recall that kernel size and “initrd” size was smaller in those days.
