[OpenAFS-port-darwin] mount_afs q's

[David Botsch]

Ok. That makes sense.

Interesting, tho, that after mounting root.afs, fs lsm /Network/afs 
tells me that /Network/afs is not a mount point (which is why I didn't 
think I was looking at root.afs).

On 2003.11.04 13:33 Alexei Kosut wrote:
> On Nov 4, 2003, at 10:25 AM, David Botsch wrote:
>> Then, went to figure out how to use mount_afs to mount afs (we have 
>> -dynroot -afsdb -fakestat also as afsd options). Anyway, I ended up 
>> using:
>> mount_afs root.afs /Network/afs
>> which did work (yay!)
>> 
>> However, I noticed that a lot of cells not listed in my CellServDB 
>> (though, are in a file called CellServDB.master) were suddenly in 
>> /afs
>> Also, a "cell" for our dns name (ccmr.cornell.edu) was listed whose 
>> contents seem to be the same as our cell, which is msc.cornell.edu 
>> (probably afsdb weirdness).
> 
> This is because "root.afs" is not the same as the root of AFS.  
> Confusing, huh?  You normally use -dynroot, which generates its own 
> AFS root.  What you're seeing when you mount "root.afs" is the 
> contents of the actual "root.afs" volume on your local AFS cell.  
> This is what you'd see if you didn't use -dynroot.
> 
> So this is expected behavior.  If you want to mount what you would 
> normally see in /afs, use:
> 
> mount_afs "" /Network/afs
> 
> The empty string parameter to the mount command tells OpenAFS to use 
> its regular root, which in your case is the dynamically-generated one.
> 
> Alexei

-- 
********************************
David William Botsch
Consultant/Advisor II
CCMR Computing Facility
dwb7@ccmr.cornell.edu
********************************