[nsp] Strange IOS behaviour

[Tomas Daniska]

well, not in this cause - just beacuse ...

a) these particularly are sub-e1 channel-groups (which imho is far from a pos pipe)
b) i believe the snapshots from the 'sh int' commands were not that far apart (seconds? maybe minutes?)


and then - this also is not a case of printing signed int as unsigned - as someone mentioned before (the values should be around 2^31 instead of 2^32 as seen in the snapshots) ... it's somewhat more messed in the counter code

--

deejay  

> -----Original Message-----
> From: Warren Kumari [mailto:warren at kumari.net] 
> Sent: 9. januára 2004 15:50
> To: cisco-nsp at puck.nether.net
> Subject: Re: [nsp] Strange IOS behaviour
> 
> Hi,
> 
> Actually, I saw this happening on packet counter, not error counter, 
> but yes, you can get that many errors (esp if you have lost 
> clocking / 
> syncronization)
> 
> If it has been happening for 90 days, you only need 552 per second 
> (2^32 / (24*60*60*90) = 552) which is not unreasonable, esp 
> on PoS. The 
> error counters are not per interval, they get cleared by 
> "clear count" 
> or reloads, etc.
> Warren.
> 
> On Jan 9, 2004, at 8:54 AM, Tomas Daniska wrote:
> 
> > 2^32 errors in one interval?
> >
> >
> > try doing some math to find out if it's possible...
> >
> > --
> >
> > deejay
> >
> >> -----Original Message-----
> >> From: Michael Costello [mailto:mikeco at corp.ptd.net]
> >> Sent: 9. januára 2004 14:32
> >> To: cisco-nsp at puck.nether.net
> >> Subject: RE: [nsp] Strange IOS behaviour
> >>
> >>> Yup, I have seen that happen a few times. The counters get
> >> up to 2^32
> >>> (4294967296) and then start counting down again.
> >>
> >> If this is true, the larger problem may be that you have more than
> >> 4294967296 CRC errors on your serial lines.