FAI Stable release candaidate 3.4.3~rc1

[Michael Tautschnig]

Hi Peter,

Thanks a lot for providing all the information!!

> Hi Michael,
> 
> Michael Tautschnig wrote:
> >Thanks a lot for debugging this issue and providing a patch.  Could you please
> >attach a config where you experienced these issues? I do acknowledge the fact
> 
> it was a simple config like:
> 
> disk_config sda preserve_reinstall:1,2,3
> 
> primary         /var    100     xfs     defaults
> primary         swap    256     swap    sw
> primary         /scratch        100-    xfs     defaults
> 

Ok, so at least my assumption that such errors only appear once you preserve a
swap partition are confirmed. Probabably this is also the reason that it hasn't
been reported before.

> 
> >that there is a bug, but I believe the bug should be fixed in some other way;
> >did you also test your patched version? I'm not sure whether parted will
> 
> of course I did.
> 

Was just curious, because ...

> >actually accept linux-swap(v1) as specification. I think we should be rewriting
> >linux-swap(v1) to swap or linux-swap in Volumes.pm.
> 
> parted does seem to use linux-swap(v1), if I do "parted -s /dev/sda print":
> 
> Model: ATA QEMU HARDDISK (scsi)
> Disk /dev/sda: 10.7GB
> Sector size (logical/physical): 512B/512B
> Partition Table: msdos
> 
> Number  Start   End     Size    Type     File system     Flags
>  1      32.3kB  98.7MB  98.7MB  primary  xfs
>  2      98.7MB  362MB   263MB   primary  linux-swap(v1)
>  3      362MB   10.7GB  10.4GB  primary  xfs
> 

... I was kind of aware of the fact that linux-swap(v1) was shown by parted, but
its documentation does not state that it would also be acceptable as input to
mkpart and frieds. But of course I hadn't thought of the fact that my code would
try to rebuild a partition as shown by parted print, and hence use the same
string, including the (v1). 

Peter, would you be available for testing a patch, or even one of the
experimental versions? I'll try to re-review your patch as soon as possible and
might commit a somewhat different version.  Yet, I don't see this as critical as
it's just about preserving swap partitions, which does seem to be a very rare
scenario.

Best,
Michael

-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 833 bytes
Desc: not available
Url : http://lists.uni-koeln.de/pipermail/linux-fai-devel/attachments/20101005/e6bc6536/attachment.bin