[j-nsp] TCAM full on EX8200?

Jeff Wheeler jsw at inconcepts.biz
Sun Oct 23 16:32:38 EDT 2011


On Sat, Oct 22, 2011 at 12:19 PM, Pavel Lunin <plunin at senetsy.ru> wrote:
> As far as I understand, it's not really correct to compare difficulty of
> these two operations, since they are performed by two different units inside

"I don't think you should compare difficulty of building a birdhouse
to building a doghouse, because a doghouse requires more materials."
You could use that argument over and over again, and as long as people
actually believed it made sense, they may agree with you.  Except it
doesn't.  Some things are harder than others.

> router. Hashing ALU's life is not a peace of cake either. Say, EX series PFE
> use only 6-bit seeds to construct hash on them. In case you want to push a
> whole 20-bit label to the hash seed, I'm afraid, you'll need more bits in
> ALU registers, more cycles or something else.

Do you realize that the source and destination IP address, TCP ports,
MAC addresses, and so on, are all larger than 20 bits?  If the thing
can figure out how to hash on those parameters, it could also figure
out how to hash on labels.

> I've heard (please correct me if I'm wrong), that the $1 per bucket ASICs,
> used in switches, are VLIW, which is hard to reprogram. While the more
> expensive ones, custom developed for routers, are rather sort of more
> flexible tiny MIMD computers with asynchronously working units inside.

To focus just on the instruction format is a bit myopic, but again, if
a chip can do Ethernet switching and IPv4 / IPv6 routing, it can also
easily be made to do label switching.  Push/pop/swap is no more
difficult than dealing with a mixture of native and 802.1q ports.

The hardware side of things can seem very complex until you understand
it thoroughly.  Then you realize what Richard keeps saying: the secret
sauce is really the control-plane software.  Hardware is basically a
commodity now.

-- 
Jeff S Wheeler <jsw at inconcepts.biz>
Sr Network Operator  /  Innovative Network Concepts



More information about the juniper-nsp mailing list