[22:56] <bbucommander> Hello all.  Not sure if this the right place to ask, but I am trying to run a shell script to mount my RAID 5 disk on boot and am wondering how I can use Upstart to accomplish this.  I followed the basic Upstart tutorials and tried placing a source reference to the script in pre-start and post-start sections of several files, but so far no luck.
[22:57] <bbucommander> I am using Ubuntu 10.10 with latest updates.  My RAID 5 is fakeraid using Intel Matrix Raid tech.
[22:58] <ion> In Ubuntu the udev stuff should handle that. If it doesn’t you probably should report a bug against whatever package provides the required stuff for the fakeraid in question.
[22:59] <bbucommander> The problem is that my RAID 5 partition is larger than 2 TB, so I have to use GPT for my partition table.  Since dmraid does not support GPT out of the box, I have to run a script calling kpartx to handle the task.
[23:01] <ion> kpartx probably should come with the necessary udev stuff to automate that then.
[23:02] <bbucommander> Is there a systematic way to check for udev automation?
[23:02] <bbucommander> Or an easy way? :-)
[23:03] <ion> Look in /lib/udev/rules.d
[23:03] <ion> Perhaps dpkg -L kpartx | fgrep udev/rules.d
[23:04] <bbucommander> Sure enough!  I found: /lib/udev/rules.d/95-kpartx.rules
[23:04] <JanC> bbucommander: do you have multiple partitions on the RAID 5 disk ?
[23:05] <bbucommander> No, just one.
[23:05] <JanC> then why use a partition table at all?  ☺
[23:06] <bbucommander> Could I still read the disk from Windows without one?
[23:06] <JanC> maybe not
[23:06] <JanC> no idea
[23:07] <bbucommander> I will look into it.  Good idea, hadn't thought of not using a partition table whatsoever.
[23:07] <JanC> I think Windows can mount USB flash sticks without a partition table, so in theory it should be possible
[23:10] <bbucommander> Forgive my ignorance of udev, but are all udev rules executed every boot?  Is there way to check if the kpartx rule is firing?
[23:15] <bbucommander> Ah, never mind.  I stumbled upon an Ubuntu bug report that essentially cleared up the issue.  The udev rule for kpartx is in fact buggy and needs repair.  Thanks ion and JanC for your help!