n Fri, Oct 16, 2015 at 9:03 AM, Phil Mayers <[email protected]> wrote:
> On 16/10/15 13:58, Mark Tinka wrote: > >> >> >> On 16/Oct/15 14:26, Phil Mayers wrote: >> >> >>> >>> In hindsight, I would not buy the N7k for the role we're using them in >>> for exactly this reason - nothing related to hardware or NX-OS, and >>> entirely related to BU focus and ability to get a TAC engineer with >>> both product knowledge and functional area knowledge. >>> >> >> I generally stick to whatever the initial design goal of the platform >> was, despite any morphing that may happen later on. >> > > I cast a slightly broader net - I'll use 6500 for routing - but I think > we're making points on the same spectrum, which is that a platform having a > capability doesn't necessarily make it a good choice ;o) > > If "real routers" were more reasonably priced at decent throughputs, I > might be a bit stricter in my choices, but working in an environment > without revenue reclaim, cost is definitely on my mind. > > _______________________________________________ > cisco-nsp mailing list [email protected] > https://puck.nether.net/mailman/listinfo/cisco-nsp > archive at http://puck.nether.net/pipermail/cisco-nsp/ > Sage advice. If I had my choice, we would use ASR9K for most routing functions. We get away with C6K and C6.8K due to the time-tested code base. N7K in contrast has been "challenging". Be very cautious when evaluating platforms for (I)SP functions. Make sure the BU understands and will support what you intend to do. -- Tim:> _______________________________________________ cisco-nsp mailing list [email protected] https://puck.nether.net/mailman/listinfo/cisco-nsp archive at http://puck.nether.net/pipermail/cisco-nsp/
