[c-nsp] "show platform hardware capacity" shows sup720-3b FIBtableat 98%

[Jared Mauch]

On Thu, Oct 19, 2006 at 06:27:07PM -0400, Richard A Steenbergen wrote:
> On Thu, Oct 19, 2006 at 04:17:01PM -0400, Jared Mauch wrote:
> > 	Then eventually the box will reload with
> > 
> > *Oct 17 17:56:47.517: %SYS-SP-5-RELOAD: Reload requested Reload Reason: FIB Protocol Allocation mismatch.
> > 
> > 	on the console.
> 
> Wierd, never seen that one.
> 
> > 	very lame imho, but it's a one time hit so not too bad.
> 
> Unfortunately that "one time hit" is way too long. Improving the boot time 
> should be a serious Cisco goal, and everyone who has influence should 
> request it. If they could have found a way to split the SP and RP images 
> while still achieve the native "effect", and have them both boot 
> simultaniously, they could have seriously reduced the load time AND 
> avoided the whole "images are too big for my flash" issue. Not to mention 
> that it seems the stock internal 64mb flashes load twice as fast (2MB/s) 
> as any compact flash (1MB/s), external or internal.
> 
> Stuff happens to the best of platforms, but a box crapping out is a lot 
> less painful when it is back up and working in under 30 secs, compared to 
> 5+ minutes. This is still one of Foundry's biggest selling points in my 
> mind. :)
> 
> I don't suppose any of this has been improved for the RSP720/MSFC4?

	The modular software is a big win for this.  I would
suggest refusing to purchase any platforms from cisco that
don't include this fancy feature called "protected memory".

	The legacy issues are getting to be too much of a liability IMHO.

	I've tried to explain the whole boot time issue to folks they
seem to not care that foundry is wiping the floor with them.  All I
hear is "redundant processors" which dosen't solve the true issue.  They've
all missed the mark (many vendors).

	- Jared

-- 
Jared Mauch  | pgp key available via finger from jared at puck.nether.net
clue++;      | http://puck.nether.net/~jared/  My statements are only mine.