[linux-lvm] Resising underlying array

[Georges Giralt]

Le 03/07/2013 20:54, matthew patton a écrit :
>
>> 200 MB for /boot was sensible years ago but now is a joke.
> since when? All you need is 2 kernels. Have Linux 3.x kernels gotten out of hand? I only use RHEL/Centos5 and 6 and I've never even come close to running out.
>
>> reduce the PV size without harm. But I'm unable to decide on a workflow and
>> loosing data or downtime are not acceptable options...
> then don't bother.
>
>> So I'm begging your help on a suitable workflow to make this happen.
>
> A. tell the customer "too bad" buy newer/bigger hard drives already and build new layout on new disks and DD/RSYNC data to new layout
>
> Or if you like to play with fire:
> * reduce filesystem size
> * lvresize down to match
> * resize MD member device (eg. http://webapp5.rrz.uni-hamburg.de/SuSe-Dokumentation/manual/sles-manuals_en/manual/raidresize.html)
> * create partition out of newly available slack space (use kpartx and you should avoid a reboot)
> * pvcreate on new partition, add to VG and grow LV into new space
> * resize filesystem
>
>
> _______________________________________________
> linux-lvm mailing list
> linux-lvm at redhat.com
> https://www.redhat.com/mailman/listinfo/linux-lvm
> read the LVM HOW-TO at http://tldp.org/HOWTO/LVM-HOWTO/
>
Hello Matthew,
Actually my volume group has a big lot of free PE. (the 320 Gb disk are 
only used for the OS and a couple of administrative accounts for the 
server, so even the /home is small.... )
So I think I could avoid the resizing of the file systems.... All data 
is stored elsewhere.
This is why I thought I could do it online without too much risk. Of 
course I will not do this on Friday ;-)

-- 
If the only tool you have is a hammer, you tend to see every problem as a nail.
                 Abraham Maslow
A British variant :
Any tool can serve as a hammer but a screwdriver makes the best chisel.