On 2013-01-28, at 11:35 AM, Saku Ytti <[email protected]> wrote: > On (2013-01-28 11:07 -0500), Jason Lixfeld wrote: > >>> We find that a planned outage, with advance customer warning, works >>> just fine. YMMV. >> >> This is 2013. If this was 1998, I'd agree with your disagreement. > > In which platform does ISSU work reliably? At least we and I know several > others have banned ISSU in their MX routers as it is source of many dodgy > issues. > In CAT4500 switches I hear it works quite well. > > But even when it does work, ISSU often means seconds of outage, this is not > something we can do without maintenance window announcement anyhow, so it's > rather unclear what exactly are the benefits. > If it would be sub 50ms and it would work reliably, I'd love to use it.
ISSU is never something I'd do in the middle of the day; always in a window. That said, sure, sub 50ms would be amazing. Until such a day, my customers and I would gladly accept many seconds of outage (ISSU), more frequently, as opposed to many minutes of outage less frequently (reboot). _______________________________________________ cisco-nsp mailing list [email protected] https://puck.nether.net/mailman/listinfo/cisco-nsp archive at http://puck.nether.net/pipermail/cisco-nsp/
