[nsp] Cat4000 Oddities

[Streiner, Justin]

On Fri, 24 Oct 2003, Darren Smith wrote:

> Seems a few weeks ago when I was on holiday our core Cisco 4006 (has 1 x 6
> port Gig-E & 1 x 48 port FE - ver 6.3.3) started putting random errors in
> the logging buffer ->
>
> %SYS-4-P2_WARN: 1/Blocked queue on gigaport 4, ( 19 : 894 )
>
> %SYS-4-P2_WARN: 1/Blocked queue (rx) on ports 3/1 3/2 3/3 3/4 3/6 3/5 3/7
> 3/8
>
> Every now and again it also adds:
>
> %SYS-4-P2_WARN: 1/Astro(3/1) - timeout is persisting    <- (Astro!?! Heh)
>
> Anyway, Shortly after the 'timeout is persisting' it appears interfaces are
> resetting but no real info is logged.

I ran into a similar issue on a Cat4000 a few months ago.  From what I
remember, the Astro errors are related to a chipset on the supervisor.
That chipset (again, from memory) controls communication between ports on
the blades with the Sup.  I saw similar behavior - connectivity through
the switch would occasionally work, but for he most part was nonexistent.

There are three of these chipsets on the Sup and in the case of a 48 port
10/100 meg Ethernet blade, each set handles communication for a bundle of
16 ports.

The resolution ended up being a brand-new switch, port blade and all.
Hope you have a spare, or Smartnet w/ hardware replacement.

If you're interested / need more info, I'll dig some notes out of the TAC
case I opened on this issue.

jms