«

Anyone ever seen this? Two 1750s; serial 0/0 are CSU/DSU WICs
connected over a leased T1 line. I do a 'show log' and see that the
line goes down 2hrs, and almost exactly 2hrs later, I see it comes up.
And stays up 2hrs, goes down. Down 2hrs, back up. Over and over. When
its up everything works fine, no errors on the interface at all. When
its down, a 'show int s 0/0' shows that the line is up, protocol down.

Are there any settings that I could have messed up that would cause
this? I have a *really* basic config on these. They have the Firewall
image, but I haven't even enabled it yet. Just a couple of routers
using HDLC and TCP/IP. Using static default routes. The T1 uses ESF
and B8ZS.

I suspect the Telco, but since they claim they can loop the
smartjacks, I'm SOL with them.

Appreciate any advice you can offer.

Could well be a clocking problem...the interface is cycling in and out of
being in sync with the provider.

Check whether your provider is supposed to provide clocking or if you are
supposed to clock internally on the type of service you have.

With line clocking ("clock source line" under the controller) both routers
get a signal from the telco.

With internal clocking ("clock source internal") only one of the routers has
to be configured this way as it will then transmit the clock signal on the
line.

By the way not sure if your WIC has a T1 controller. If not the config may
be under the serial interface itself and the commands may differ a little
(transmit-clock-source? from memory)



"Ivan" <> wrote in message
news: om...
> Anyone ever seen this? Two 1750s; serial 0/0 are CSU/DSU WICs
> connected over a leased T1 line. I do a 'show log' and see that the
> line goes down 2hrs, and almost exactly 2hrs later, I see it comes up.
> And stays up 2hrs, goes down. Down 2hrs, back up. Over and over. When
> its up everything works fine, no errors on the interface at all. When
> its down, a 'show int s 0/0' shows that the line is up, protocol down.
>
> Are there any settings that I could have messed up that would cause
> this? I have a *really* basic config on these. They have the Firewall
> image, but I haven't even enabled it yet. Just a couple of routers
> using HDLC and TCP/IP. Using static default routes. The T1 uses ESF
> and B8ZS.
>
> I suspect the Telco, but since they claim they can loop the
> smartjacks, I'm SOL with them.
>
> Appreciate any advice you can offer.

(Ivan) wrote in message news:<. com>...
> Anyone ever seen this? Two 1750s; serial 0/0 are CSU/DSU WICs
> connected over a leased T1 line. I do a 'show log' and see that the
> line goes down 2hrs, and almost exactly 2hrs later, I see it comes up.
> And stays up 2hrs, goes down. Down 2hrs, back up. Over and over. When
> its up everything works fine, no errors on the interface at all. When
> its down, a 'show int s 0/0' shows that the line is up, protocol down.
>
> Are there any settings that I could have messed up that would cause
> this? I have a *really* basic config on these. They have the Firewall
> image, but I haven't even enabled it yet. Just a couple of routers
> using HDLC and TCP/IP. Using static default routes. The T1 uses ESF
> and B8ZS.
>
> I suspect the Telco, but since they claim they can loop the
> smartjacks, I'm SOL with them.
>
> Appreciate any advice you can offer.

You may want to look at the clocking. If your deriving clock from the
circuit, it will be harder to track down a clocking problem, but it
can be done (using a T-Berd or similar diagnostic device). Another
possibility would be some kind of memory leak issue, but I seriously
doubt that you would see the kind of regular interface cycling that
your experiencing with a memory leak.
HTH, David

"Ben" <> wrote
> Could well be a clocking problem...the interface is cycling in and out of
> being in sync with the provider.
>
> "Ivan" <> wrote in message
> news: om...

> > Anyone ever seen this? Two 1750s; serial 0/0 are CSU/DSU WICs
> > connected over a leased T1 line. I do a 'show log' and see that the
> > line goes down 2hrs, and almost exactly 2hrs later, I see it comes up.
> > And stays up 2hrs, goes down. Down 2hrs, back up. Over and over. When
> > its up everything works fine, no errors on the interface at all. When
> > its down, a 'show int s 0/0' shows that the line is up, protocol down.

I am not complaining about Ben's response AT ALL. Indeed it
seems a fine plan.

This however seems intriguing.

2 x 1.5 Mhz clocks drifting by 1 cycle every 4 hours.

That is a rate of 1 second in 700 years.
Hmmmm - when did Cisco start fitting Atomic clocks to the kit?

Oh no!
They seem to do even better, http://whyfiles.org/078time/2.html -
"To put it another way, this clock will stay within one second of
true time for 6 million years"

Well it seems it might be true.
"The NIST program on chip-scale atomic clocks" seems to have EXACTLY
the right accuracy

Thanks for the replies. I meant to include the fact that I am doing
the internal clocking on one of the routers and that I had double
checked that.

I ended up calling the Telco on their bluff and their dire warnings
that I would be charged if they had to come out and the problem was on
my end. They came out, fiddled around, had someone at the CO fiddle
around, and when they were done the line came up and has stayed up (so
far). They never really came out and said how they fixed it. I'm not
sure they really know.

That's been pretty much the characteristic of the Telcos.
They don't want to admit it is their problem and they don't want to
make a trip out when they can scare you into thinking that you might
be charged for the trip.

More times than not, it is their problem. Rarely is their a
wiring problem on your end. If it is a customer equipment problem,
then you should have checked all of your equipment before calling
them.

I've actually had times when they said there was no problem on
the line but miraculously everything that wasn't working before they
looked at it is now working. Coincidence? Hardly.

I think that they think they will be shot if something is
wrong with their equipment.



Fred

In article <>,
says...
> That's been pretty much the characteristic of the Telcos.
> They don't want to admit it is their problem and they don't want to
> make a trip out when they can scare you into thinking that you might
> be charged for the trip.
>
> More times than not, it is their problem. Rarely is their a
> wiring problem on your end. If it is a customer equipment problem,
> then you should have checked all of your equipment before calling
> them.
>
> I've actually had times when they said there was no problem on
> the line but miraculously everything that wasn't working before they
> looked at it is now working. Coincidence? Hardly.

*snort* you too? "honestly, no-one touched ANYTHING for your line
and now it's working? it's a miracle!!!"

Isn't that a famous quote from "The Shawshank Redemption"? I
think it's what the warden said when he didn't know how Andy had
escaped from his cell.

Regards,



Fred

In article <>,
says...
> Isn't that a famous quote from "The Shawshank Redemption"? I
> think it's what the warden said when he didn't know how Andy had
> escaped from his cell.

I was always amused by trying to guess how long it would be after
opening a ticket with Bell Atlantic before the T1 (which had been down
for hours) would magically heal itself...

I can see that we've both had the same experiences.

I used to be Telco support (among my other functions) for
SkyTel Technical Operations. We used data and dialup lines from all
over the country. I interfaced with many different telephone
companies all over the country.

Sometimes they'd try to snow me. Sometimes I escalated
through them internally. On rare occasion, I called whichever state's
Public Service Commission was appropriate. But, I always got it
resolved when no one else could. But if we had to rely on those
Telcos, a lot of problems wouldn't have gotten resolved by itself.

When I worked for MCI, one of my later specialties was
interfacing with troubled accounts. When they sent me over to fix
someone's WATS or data lines, it was generally because those customers
were ready to eat the next representative of our company alive.
Didn't bother me, though. I always had them eating out of my hand
before I left. And I rammed heads with Telco field personnel more
times than I wish I had had to. It only took a few times where they
told me there was no problem on the line and I proved them wrong
before they had a chance to regroup. Sometimes, I fixed their
equipment myself.

After I started making them look so bad, the Telco field guys
got the word about me. After that, when I encountered them and asked
them to do something, it was 'Yes, sir' with no argument. They did
it, too.

Good to bump into someone with similar experiences.



Fred