siloconfig write incorrect silo.conf in some circumstance

Bug #177681 reported by Etienne Goyer
2
Affects Status Importance Assigned to Milestone
silo (Ubuntu)
Confirmed
Undecided
Unassigned

Bug Description

Binary package hint: silo

siloconfig is a Perl script being called by the silo postinst to write a silo.conf when the installation is interactive. Under some circumstance, it could write an incorrect silo.conf.

siloconfig derive the value of the "partition" directive of silo.conf by truncating the device name where the root filesystem reside, according to /etc/fstab. This could lead to an incorrect value for the partition directive if the partition scheme do not follow the typical partitioning scheme. As an example, if the root filesystem reside on the software RAID volume with device name /dev/md3, "partition=3" will be written to silo.conf, which would be wrong as the third partition is the "Whole disk" one.

Revision history for this message
Daniel T Chen (crimsun) wrote :

Is this symptom still reproducible in 8.10?

Changed in silo:
status: New → Incomplete
Revision history for this message
Etienne Goyer (etienne-goyer-outlands) wrote :

Unfortunately, I cannot say as I do not have physical access to a SPARC machine anymore to test. However, checking the changelog in Intrepid, there is no mention of any change in siloconfig. In fact, debdiff'ing the latest silo 1.4.14-0ubuntu1 with 1.4.13-1ubuntu5, there have not been any changes in siloconfig that I can see. So I would say the bug is still there.

Revision history for this message
Michael Casadevall (mcasadevall) wrote :

Just ran into this tonight. Still there. The problem is siloconfig (and silo in general) do not like UUIDs.

Changed in silo (Ubuntu):
status: Incomplete → Confirmed
To post a comment you must log in.
This report contains Public information  
Everyone can see this information.

Other bug subscribers

Remote bug watches

Bug watches keep track of this bug in other bug trackers.