N7K F1 cards vs M1 cards

[bruno.fernandes]

Hello,

I'm trying to understand the correct positioning of the F1 cards in terms of positioning, from all the info that I have read I come into the conclusion that F1 cards doesn't do L3 functions, unless you use the resources from an M1 cards, this brings me several questions an possibilities:

1. If the F1 cars use L3 resources from M1 does this means that I have to make some configuration to enable this feature ? Or it's automatically enabled ?

2. I still don't understand where this F1 cards are suited for ? Just have looked into the documentation ...and nothing just the usual "Low latency cards, no L3 resources.....no info on positioning in the network or at least I didn't find it)

3. In your opinion are they suitable for Aggregation layer (uplinks from access switches to core) ?

4. If they use L3 resources from M1 cards, does this means they will need to make a new L3 lookup for each new packet that ingress the F1 card ?

5. And how about redundancy, If I use the resources from and M1 card and this card fails... ?

6. And regarding performance ? From What I understand If need and L3 lookup in and F1 card the packet must travel trough the fabric to the M1 Card, take the decision and travel again trough the fabric and leaves the switch completing the 3rd stage.

Kind regards,

Bruno Fernandes

[Muhammed Marakkoottathl]

Hi Bruno,

Please see the answer to your queries below.

Q 1. If the F1 card use L3 resources from M1 does this means that I have  to make some configuration to enable this feature ? Or it's  automatically enabled ?

Ans :F1 modules do not natively provide Layer 3 switching and it cannot inter-VLAN route on their own However, one or more M1/M1-XL modules can provide “proxy” Layer 3 services.

Proxy L3 forwarding enabled by default when VDC in mixed- module mode

Packets destined to router MAC forwarded to M1/M1-XL modules for Layer 3 via internal “Router Port-Channel” (RPC) and

Specific module selected based on EtherChannel hash function

Q 2: Where this F1 cards are suited for ? Just have  looked into the documentation ...and nothing just the usual "Low latency  cards, no L3 resources.....no info on positioning in the network or at  least I didn't find it)

Ans: F1 card you can use when you require high density L2 performance.

Q 3: In your opinion are they suitable for Aggregation layer (uplinks from access switches to core) ?

Ans: Typical deployment model will be at aggregation layer, with F1 L2 ports facing access/downstream, and M1 L3 ports facing core/upstream.

Q 4. If F1 use L3 resources from M1 cards, does this means they will  need to make a new L3 lookup for each new packet that ingress the F1  card ?

Ans: it works in this way , traffic will flow from F1 module to the backplane, than enter into M1 module, perform the routing and get out to backplane again, the M1 module / forwarding engine is used is based on the L2 Portchannel hash result ( it is something that is called a "router portchannel which is an internal switch construct)

Q5: And how about redundancy, If I use the resources from and M1 card and this card fails... ?

Ans: The redundancy is depended on how many M1 module you have in the chassis.  If you have more than 1 M1 module, the load-balance happense across these M1     module

Hope this will help.