[comp.dcom.lans] wrap vs. bypass

goodloe@xenon.UUCP (Tony Goodloe) (12/22/88)

with regards to FDDI: what is the distinction between "wrapping" and
"bypassing"? thanks in advance.

kwe@bu-cs.BU.EDU (kwe@bu-it.bu.edu (Kent W. England)) (12/24/88)

In article <140@xenon.UUCP> goodloe@xenon.UUCP (Tony Goodloe) writes:
>with regards to FDDI: what is the distinction between "wrapping" and
>"bypassing"? thanks in advance.

	I am familiar with the terms from Pronet-80 and I assume they
carry over intact to FDDI, but wrapping and bypassing are descriptive
terms.
	"Wrapping" is when a dual link goes down (infamous backhoe
cut) and the nodes on each end of the failed link detect the loss of
primary and secondary rings and loss of neighbor and then they "wrap"
the primary and secondary rings together, connecting the two
counter-rotating rings together into a single ring in the shape of a
"C".  Full connectivity restored.
	"Bypassing" is when the FDDI node detects a problem with the
local ring or host and disconnects the local network from the
backbone.  It bypasses a failed or inactive host/ring.  This maintains
the two counter-rotating rings intact and fully operational, since the
bypassing node is repeating frames around the ring and is fully
participating in the dual ring management, etc.  It's a necessary
token ring fault management technique.

	Kent England, Boston University