FAI Stable release 3.4.0

[Michael Tautschnig]

Hi Julien,

[...]

> 
> The change to Exec.pm in setup-storage to add a catch-all on command
> errors breaks RAID setups, see the attached format.log (I can send the
> config file if needed).
> 
> Parted exits with this error:
> 
>   Warning: WARNING: the kernel failed to re-read the partition table on
>   /dev/sdb (Device or resource busy).  As a result, it may not reflect all
>   of your changes until after reboot.
> 
> As /dev/md0 is started before this parted command is executed, both
> disks are now busy and the kernel can't re-read the partition table.
> 

Would it be possible to verify that the very same problem appears if you use the
experimental FAI packages? I think there shouldn't be any difference, but just
wanted to make sure that a future patch really fixes this issue.

> It is probably not possible to let this particular error through as it's
> too generic and could be the result of other more critical commands
> earlier on.
> 
> Unfortunately, parted doesn't seem to have an option to NOT re-read the
> partition table after making changes :/
> 
> I think the best we can do is to add || true to this particular command
> in set_partition_type_phys_dev(), at least for now. Anything else will
> probably be too invasive for 3.4 at this stage...
> 

[...]

I'd like to fix this by enforcing an order of commands where all partition-table
related commands appear before any array on that disk is started. The patch that
is supposed to achieve this is ready, just waiting for the above
cross-validation before committing it to experimental.

Thanks a lot,
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/20100816/c1b783e8/attachment.bin