We had a similar situation for 3 or 4 days, when suddenly it burst into
life. Frantic phone calls later its discovered someone had turned off the
Vj-Comp and STAC, which was supposed to have been turned off in the first
place.
Ensure there is *NO* STAC or ANY compression turned on on the Cisco. This
*ALSO* means VJ-Header Compression.
Once that was done, the connection went like a dream.
At 02:50 8/08/97 +0200, Willem Jan Withagen wrote:
>
>HI
>
>We've had a lot of trouble upgrading one of our links to 2Mbit today,
>which started when our dutch telco messed up the CPU board in our
>multiplexer. Which took them nearly 8 hours to repair, and giving us only
>a 1920Kbit stream/V.35.
>
>Now did I think that that would not matter since this is SYNC traffic and
>it is self-clocking. (A PM does take a 144Kbit stream from one of our
>baseband modems). But our experience seems to indicate otherwise.
>The ISP on the other side has a CISCO, and the connection work flawless at
>1 Mbit.
>
>What we noticed is that at 2 Mbit the irx spews out a lot of
>LCP_REQUEST_CONFIG
>And only once of about every 15-20 something is received back from the
>CISCO. What the CISCO sends is correctly recognised by the IRX as being
>again an LCP_REQUEST_CONFIG.
>
>Once in a while the IRX tell us that it is reset it's DMA???
>We got lots of aborts on the serial port, but few CRC's and or Frame errors.
>
>My conclusions is that both systems don't realy understand one another
>probably 'cause the very basic layer is not working correctly.
>But I'm to understand that the IRX should be able to do atleast 2Mbit on
>1 port without getting into trouble?
>
> - So what are we getting?
> - Is the system overloaded?
> - Doesn't it take 1920K, which BTW is shown as 2048K external clocked
> with the newest COMOS software release?
> - The Telco person mumbled something that the signal had it perhaps
> clock-flanks reversed, and he had to switch his tester.
> Could and IRX be succeptible to this?
>
>Problem is that testing this again is not very easy, not unless we get our
>second feed up. But that's again 2Mbit which we were going to do with IRX's
>and BGP. Testing might occur during off hours: 3 at night or so.
>
>We've now used a CISCO 2501 which does not see any sort of problems
>and is taking the traffic oke but it will not do BGP as we would like it to
>do: 2 full feeds.
>
>Does anybody have suggestions on how to tackle this?
>Other that buy more CISCO's?
>
>Thx,
> Willem Jan
>
>--
>Internet Access Eindhoven BV., voice: +31-40-2468500, data: +31-40-2439436
>P.O. 928, 5600 AX Eindhoven, The Netherlands
>Full Internet connectivity for only fl 12.95 a month.
>Call now, and login as 'new'.
>
>
-- ************************************************************************* *Adrian Carter Email: adrian@apic.net * *Network Administrator URL: http://www.apic.net/ * *The Asia Pacific Internet Company Pty Ltd Autoresp: info@apic.net * * All programmers are playwrights and all computers are lousy actors. * * Phone: (+612) 9419-5333 Fax: (+612) 9419-5155 * *************************************************************************