slowwness at boot (hours)

[Matthew Godycki]

On March 31, 2006 11:50 am, John Van Ostrand wrote: 
> On Fri, 2006-03-31 at 10:24 -0500, Marc Lijour wrote: 
> > Hi 
> > 
> > I have an issue with my RAID1 system. I use Mandriva 2006 on x86_64. 
> > 
> > First the system hangs while running fsck on one partition. I solved this 
> > by booting on a live cd and running fsck from there. I ran fsck -y 
> > /dev/sd{a,b} 5. 
> 
> Using a Journalled file systems like ext3 will alleviate an extended 
> fsck in the event of a hard shutdown. 
 
| My bad, I should have mentioned I use ext3. 
| Precisely the problem started with a hang while the system was "recovering the  
| journal". 
 
> > From the live cd, I also ran mdadm --examine  and it tells me the 
> > drives are in sync. 
> > 
> > Now, I reboot from the hard drive and I get one array which is missing a 
> > drive. But worse, the boot process is extremely slow (it can take hours 
> > to fire one daemon). I have also a couple of bad looking reports such as 
> > /dev/tty0 does not exist, and some other config files missing. Would you 
> > know what is causing this? Is the data corrupted? Can I get it back? 
> 
> Check your DNS. If a service is reliant on DNS it can take some time to 
> come up, however I would expected minutes per daemon, not an hour. 
 
| All services are slow. My first guess was that the raid arrays are syncing,  
| but I am not sure anymore. Because I don't have a prompt it is hard to say. 

 
 I'm going to guess that you're experiencing hardware failure.  
  The only times I've ever seen slow reading disk drives and 
  otherwise erratic behaviour like you describe has been when 
  I've had hard drives fail.
 
 -M
  
 


--
The Toronto Linux Users Group.      Meetings: http://tlug.ss.org
TLUG requests: Linux topics, No HTML, wrap text below 80 columns
How to UNSUBSCRIBE: http://tlug.ss.org/subscribe.shtml