[OpenAFS-devel] mountpoints on linux

Gary Buhrmaster gary.buhrmaster@gmail.com
Fri, 29 Jun 2012 21:55:38 +0000


On Fri, Jun 29, 2012 at 9:16 PM, Andrew Deason <adeason@sinenomine.net> wro=
te:
.....
>> =A0 =A0 =A0D. Treat every volume as a separate filesystem, like kafs doe=
s.
>> =A0 =A0 =A0 =A0 While this has some advantages, it also has some serious
>> =A0 =A0 =A0 =A0 disadvantages. =A0I also have a vague recollection of co=
ming up
>> =A0 =A0 =A0 =A0 with a reason at one point why this model is fatally fla=
wed.

I am guessing here, but at least in kernels before 2.6 the maximum
number of minor device numbers was limited to 255 (with 2.6, that
is now 2^20).  Also, all anonymous file systems share the same
major number (so the 255 was shared between afs, nfs, /proc, etc)....
Of course, you may actually have a different (and better) reason
for the fatal flaw.