Bug 148098

Summary:	partition proposal kind of nonsense
Product:	[openSUSE] SUSE Linux 10.1	Reporter:	Hans-Peter Jansen <hpj>
Component:	Installation	Assignee:	Thomas Renninger <trenn>
Status:	RESOLVED WONTFIX	QA Contact:	Klaus Kämpf <kkaempf>
Severity:	Normal
Priority:	P5 - None	CC:	hpj, suse-beta
Version:	Beta 3
Target Milestone:	---
Hardware:	i686
OS:	SUSE Other
Whiteboard:
Found By:	Beta-Customer	Services Priority:
Business Priority:		Blocker:	---
Marketing QA Status:	---	IT Deployment:	---

Description Hans-Peter Jansen 2006-02-03 19:54:20 UTC

On a system with this existing partition table:

Disk /dev/hda: 80.0 GB, 80060424192 bytes
255 heads, 63 sectors/track, 9733 cylinders
Units = cylinders of 16065 * 512 = 8225280 bytes

   Device Boot      Start         End      Blocks   Id  System
/dev/hda1               1         131     1052226   82  Linux swap / Solaris
/dev/hda2   *         132         135       32130   83  Linux
/dev/hda3             136        1011     7036470    7  HPFS/NTFS
/dev/hda4            1012        9733    70059465    f  W95 Ext'd (LBA)
/dev/hda5            1012        2924    15366141    7  HPFS/NTFS
/dev/hda6            2925        4955    16313976   83  Linux
/dev/hda7            4956        8220    26221127   83  Linux
/dev/hda8            8221        9732    12145108+  83  Linux

YaST proposes to:
destroy /dev/hda7 (25.0 GB)
create /dev/hda8 (15.0 GB) for /home
create /dev/hda9 (10.0 GB) for /

I haven't tried to accept this proposal, since it would destroy my most valuable partition on it, but accepting this surely result in some trouble, and reveals a couple of questions:

 - Does YaST get the partition ordering right?

It kills an arbitrary part., only to cut it into pieces, with /home first, and / second? 
What about the currently existing part. /dev/hda8? In order to make (some
limited) amount of sense from what YaST wants to do, it will move /dev/hda8
to /dev/hda7 (recipe for trouble), and append the two new ones to the end
(hda8 and hda9), again calling for trouble, since they aren't ascending
anymore.

 - It presents a major inconvenience, since I cannot modify this proposal to make some more sense, thus I have to start from scratch - and is a step back IMHO. Update: I found the way to select a certain part., but nevertheless, I had to add the (formerly automatically detected) win mountpoints back.

In my opinion, YaST shouldn't try to make some arbitrary decisions in such
a case: better give the user some hints about what's in those part., and let the user decide... I easily can change to tty2, mount them by hand and check
what's in, but that's not something, joe avarage user would do...

...and let's hope, that 10.1 final will have xfs FS back.. :-(

Comment 1 Thomas Renninger 2006-02-06 18:53:59 UTC

The current logic is without sniffing any information from any partition and this is good. You cannot check whether a partition is a home/usr/xy partition that wants to be reused/reformatted.

The logic is based on file system types and size of a partition, what should give you a good idea of partitioning.
You can choose the based on suggestions (or similar) option. Like that the partitioner will still recon an existing swap/NTFS/VFAT fs and mount it correctly.
You can then alter the partition table and file system types and where they should get mounted manually and rather easy.

At least you should know what is on your linux file system partitions. If you do not, you also don't want to trust your data to an automatic assumption and still go this way:
-> I easily can change to tty2