Framebuffers, plymouth, upstart and server installs.

[Mark - Syminet]

On Jan 9, 2013, at 2:48 AM, Bouchard Louis <louis.bouchard at canonical.com> wrote:
> 
> I've been puzzled by this thread for a while.  I'm booting Ubuntu
> servers all day long (mostly in VMs nowaday) and all I see is the stream
> of console messages up to the boot prompt.  Here is a small capture of
> one Precise server boot :
> 
> http://people.canonical.com/~lbouchard/precise_server_boot.ogv
> 

That's because you have a small drive on your VM.  Like I said, for desktops 
this is probably OK (Macs do same) but what if the host for those VM's has a 
10TB disk array and suffers a powercut?  A 10TB fsck can take hours., and 
what happens is that you see this line with the flashing cursor: 

Press C to cancel all checks in progress
_ 

…nothing else.  That should read with a moving progress bar, like this 
(and it was like this for over a decade until upstart arrived): 

Press C to cancel all checks in progress
[=======                                               ]  24%

…in your video at 0:17 we see: 

Press C to cancel all checks in progress
/dev/vda1: 230/124496 files (1.3% noncontiguous… etc. 

…the progress bar is missing.  It should look something like this: 

Press C to cancel all checks in progress
[========================] 100% Complete
/dev/vda1: 230/124496 files (1.3% noncontiguous… etc.  

I know this sounds nit-picky, and it's true that it isn't the end of the world - but 
consider servers with very large disks, and people (read: your boss) screaming 
the whole time for ETA's to uptime - suddenly that little bar becomes quite 
valuable.  At least we know if the server is doing something, and whether or 
not it's hung.  

Best - 

-- 
Mark