Ascend Archive
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
RE: (ASCEND) Max flooding our ethernet backbone
On Thursday, August 27, 1998 1:35 PM, Jason Nealis
[SMTP:nealis@babylon.erols.com] wrote:
>
> Any chance you could get a linux box on the segment and run tcpdump, or
> run some sort of sniffer on that physical segment?
>
> Jason Nealis
> Director Internet Operations / Network Access
> Erols Internet (An RCN Company)
>
>
> On Thu, 27 Aug 1998, Gavin Cato wrote:
>
> > Hi Everyone,
> >
> > We have 2 Max4000's in one of our POP's that feed into a bay 350T 10/100
> > switch.
> >
> > Since yesterday morning one of the Maxes is flooding our Ethernet backbone
> > in intervals of roughly 10-15 seconds (sometimes slightly longer)
> >
> > When it does it all the lights on the hub light up & stay *constant*
> >
> > My MRTG graph shows that when it happens every single machine connected to
> > the ethernet suddently shoots up by about 50-60kb/s
> >
> > If I unplug the Max during one of these spurts, it immediately goes away.
> > If I unplug it & leave it out - everything runs fine.
> >
> > The MRTG graph that I have that sums up all 16 ports on the hub into 1
> > graph shows it typically sending about 700kb/s. Typically this hub sends
> > about 60-100kb/s
> >
> > Any ideas? I'm running out of things to do. Both Maxes are running 6.0.2
> > and have been fine for serveral months without changing anything. Remember
> > only 1 is playing up.
> >
> > I upgraded it to 6.1.7 but that didn't change the situation so I
> > downgraded it back to 6.0.2
> >
> > Any clues?
> >
> > thanks
> >
> >
> > Gavin Cato
> > System Administrator
> > Geko Internet P/L
> > (02) 9439-1999
> >
> > On Wed, 26 Aug 1998 msheehan@sctcorp.com wrote:
> >
> > > Upgraded our Max 4048 from 6.1.3 to 6.1.7. It's been in
> > > production for 2 days. Just received the following fatal error
> > > and reboot after 6 day of uptime. Seem to happen while I was in
> > > the Telnet screen and went to put in the security profile 2
> > > password. Might have been a coincidence. Just wanted to pass
> > > along for those thinking of upgrading.
> > >
> > > SYSTEM IS UP: Index: 100 Load: ftk.m40 Revision: 6.1.7
> > > Date: 08/19/1998. Time: 14:43:24
> > >
> > > FATAL ERROR: Index: 1 Load: ftk.m40 Revision: 6.1.7
> > > Date: 08/26/1998. Time: 15:55:15
> > > Location: b011722c b0118778 b004c3bc b00d0718 b004cddc 00
> > > 000000
> > >
> > > SYSTEM IS UP: Index: 100 Load: ftk.m40 Revision: 6.1.7
> > > Date: 08/26/1998. Time: 15:57:26
> > >
> > > Maryann Sheehan
> > > SCT Corp.
> > > msheehan@sctcorp.com
> > >
> > >
> > > ++ Ascend Users Mailing List ++
> > > To unsubscribe: send unsubscribe to ascend-users-request@bungi.com
> > > To get FAQ'd: <http://www.nealis.net/ascend/faq>
> > >
> >
> > ++ Ascend Users Mailing List ++
> > To unsubscribe: send unsubscribe to ascend-users-request@bungi.com
> > To get FAQ'd: <http://www.nealis.net/ascend/faq>
> >
>
> ++ Ascend Users Mailing List ++
> To unsubscribe: send unsubscribe to ascend-users-request@bungi.com
> To get FAQ'd: <http://www.nealis.net/ascend/faq>
Hi Gavin and all
We've had the same problem with Max 1800's for about 2 months now at a remote
pop.
The site runs 6 max 1800's a few ciscos and some unix boxes.
We've nailed it down to arp storms. Although, we cant work out WHY its causing
it and asking ascend is having it fall on deaf ears.
Try this: ping the suspect max and keep an eye on it. Does the icmp times go
up? If so, run a package (WebBoy is a good one) that watches for network
traffic type and see if RARP requests fire up when the ethernet becomes
slogged.
5 bucks says thats the problem! :-)
Regards
adam
--------------------------------------------------
IPAX Systems
Melbourne, Australia
National Systems Manager
Ops Tel: 1800 894 894
Ops Fax: 03 - 9801-8533
Web Site http://www.ipax.com.au
--------------------------------------------------
++ Ascend Users Mailing List ++
To unsubscribe: send unsubscribe to ascend-users-request@bungi.com
To get FAQ'd: <http://www.nealis.net/ascend/faq>