Hey Gert, > (Yes, we have templates for both pre-ELS and post-ELS, but having to > figure out what is new now and how and why port-unrelated things have > to be changed was an unneccessary waste of lifetime)
Not disagreeing, but on the scale of approving new hardware, writing new templates isn't that significant OPEX cost. New platform means learning new limitations of HW, new corner cases, new way to troubleshoot, new things to monitor, it's inevitably rather huge chore. Maybe there comes a day when we don't have to be hyper-aware of the underlaying hardware in networking kit, but that certainly isn't today, in that future, this would have been relatively much larger issue. Perhaps this is more marketing problem than technical, as they are presenting it as minor upgrade, when in reality everything changed. It would be still interesting to hear what was the reason. Would the preELS map poorly to modern switching asics, is postELS more in-line on how forwarding-plane is actually programmed? Or could they have trivially supported preELS, but regardless of BRCM noticed their first attempt at switch CLI has inefficiencies they decided to remedy while changing everything anyhow. -- ++ytti _______________________________________________ cisco-nsp mailing list [email protected] https://puck.nether.net/mailman/listinfo/cisco-nsp archive at http://puck.nether.net/pipermail/cisco-nsp/
