Lustre devel - Dec 2006 - [Bug 10802] UUID collision on thin XT3 Linux nodes

Please don''t reply to lustre-devel. Instead, comment in Bugzilla by
using the following link:
https://bugzilla.lustre.org/show_bug.cgi?id=10802

           What    |Removed                     |Added
----------------------------------------------------------------------------
  Status Whiteboard|2006-12-14: updated patch   |2006-12-31: patch landed on
                   |available                   |b1_4 for Lustre 1.4.9 and
                   |                            |b1_5; updated patch
                   |                            |submitted by Cray is pending
                   |                            |inspection by CFS


Comment Tags: Status Change

2006-12-31: patch landed on b1_4 for Lustre 1.4.9 and b1_5; updated patch
submitted by Cray is pending inspection by CFS

Oleg, can you please inspect the change submitted by Nic on 2006-11-20?

Please don''t reply to lustre-devel. Instead, comment in Bugzilla by
using the following link:
https://bugzilla.lustre.org/show_bug.cgi?id=10802

           What    |Removed                     |Added
----------------------------------------------------------------------------
   Attachment #8935|review?(green@clusterfs.com)|review-
               Flag|                            |


(From update of attachment 8935)
What I think should be really done is to move liblustre versions of
get_random_bytes/generate_random_uuid 
to obdclass/prng.c as ll_* functions.

I am sure there are some bits in UUID that are meaningful. E.g. see
kernel''s
realisation:
	/* Set UUID version to 4 --- truely random generation */
	uuid_out[6] = (uuid_out[6] & 0x0F) | 0x40;
	/* Set the UUID variant to DCE */
	uuid_out[8] = (uuid_out[8] & 0x3F) | 0x80;

Not sure how much we care about this, though.