I'm futzing with things down in InfiniLand. A new race condition
has cropped up that reorders buffer mappings between sender and
receiver, and I'm not sure why, but it doesn't matter because I'm
fixing it. But in the process I had to tweak the bits on the wire
between a client and a server, only in the IB internals.
But bad things happen if you run a soon-to-be-checked-in client
against an old server:
piv057$ pvfs2-ls -la /pvfs-ib
[E 15:12:48.332732] Error: poorly formatted protocol message received.
[E 15:12:48.333152] Encoding type mismatch: received type -1073741862
when expecting 2.
Since this is a protocol-breaking change, my question to the group
is, should I bump this?:
/* update PVFS2_PROTO_MAJOR on wire protocol changes that break backwards
* compatibility (such as changing the semantics or protocol fields for an
* existing request type)
*/
#define PVFS2_PROTO_MAJOR 1
It seems to me like a big deal for what is really just a small
change in one obscure network type. I would be happy to say that
pvfs2/ib is esoteric enough that no one will be in the position of
upgrading just a client or a server independently. Your thoughts?
-- Pete
_______________________________________________
Pvfs2-developers mailing list
[email protected]
http://www.beowulf-underground.org/mailman/listinfo/pvfs2-developers
