Nabil, et al,

Do you assume network nodes as service nodes in the context of Service Chaining?
In the same level as today's middle boxes (a.k.a. L4-L7 service nodes)?

For example, if a data flow needs to traverse through nodes: CE1, FW1, PE1, P1, 
PE2, FW2, CE2, then service chain is: CE1 -> FW ->  PE1 -> P1 -> PE2-> FW2-> 
CE2. Correct?

There are protocols (e.g. IGP/BGP) to discover network layer nodes, but many of 
today's L4-L7 service functions (so called middle boxes) don't participate in 
the routing protocols for topology discovery.
Therefore, Section 3.1 (Service topology) should differentiate the two cases. 
The attributes you listed in Section 3.1 for "Service Node" are more applicable 
to L4-L7 service nodes.

Page 7, you also listed additional attributes (many of them are for TE based 
routing), mixed with MAC/FIB database size, etc. They are more for L2/L3 device.


Section 3.2.  (Monitoring Information)
Many of today's L4-L7 service functions are monitored by their own monitoring 
systems. Monitoring FW is very different than monitoring DPI, or Video 
optimization function.

The parameters you listed in Section 3.2 are more for monitoring routers, less 
for monitoring L4-L7 service functions.

Should make it clear that those parameters are more for L2/L3 devices.


Since your section 3 is to  "describes requirements and applicability for such
information, and for directing traffic through a service chain", it is be 
beneficial to show a figure to differentiate network layer nodes from other 
service functions (e.g. FW, DPI, or so called L4-L7 service functions). To make 
it easier for the targeted description.
For example, you can use the figure I have in 
https://datatracker.ietf.org/doc/draft-dunbar-sfc-legacy-l4-l7-chain-architecture/


[cid:[email protected]]


Linda



_______________________________________________
i2rs mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/i2rs

Reply via email to