Re: (PM) Difficulty using BACP between PM3 and P50's

Prater, Chuck (praterc@nventure.com)
Wed, 21 Apr 1999 21:06:24 -0700

I've successfully implemented BACP with the Pipeline 75 dialing the PM3.
Never tried PM3 dialing Pipeline 75. Set up instructions can be found by
searching Ascend on the Livingston web page. Unless you have caller ID on
your PRI you have to do a set sx directory xxx-yyy-xxxx where xxx-yyy-xxxx is
the phone number needed to be dialed by the pipeline.

Bruce Walker wrote:

> I've been struggling with a new PM3 (running 3.8.2) talking to
> Ascend P50's (old and new; I have about six of them). My test P50
> is running 6.0.13.
>
> All I'm trying to do is to have the link between the PM3 and the
> P50 add a channel as load increases and remove it when the load
> drops. Whether that requires MP, MPP, BACP or DBA, I don't care!
>
> So far I can dial up a P50 and successfully negotiate VJ, Stac LZS
> and Multilink. I can load down a single channel with no trouble for
> hours on end (eg: setting maxport=1 in the location profile on
> the PM3).
>
> But if I allow a second B channel to come up (maxport=2), what
> happens next depends on how I set things on the P50, but it always
> screws up.
>
> If I set MP encaps and leave "Sub Pers=10" (the default of 10
> seconds) on the P50, then shortly after dialing a second link, the
> PM3 receives a BAP_LINK_DROP_QUERY_REQUEST almost immediately after
> the 2nd link is up. The PM3 replies with a BAP_LINK_DROP_QUERY_RESPONSE
> and drops the 2nd channel. Then the link "freezes". No TCP traffic
> gets through. There's still one link up as seen by "show session"
> but there's no data. This is usually accompanied with a series of
> these:
>
> Received CCP_RESET_REQUEST on port S3 of 2 bytes containing:
> 0E 02 00 06 00 01
> Packet Info: Code: 0x0E, ID: 0x02, 6 bytes.
> History Number: [0x0001], No History number is present. (RFC1974)
>
> Sending CCP_RESET_ACK to port S3 of 6 bytes containing:
> 0F 02 00 06 00 01
> Packet Info: Code: 0x0F, ID: 0x02, 6 bytes.
> History Number: [0x0001], No History number is present. (RFC1974)
> [...]
>
> The other behaviour is if I set "Sub Pers=120" on the P50. I
> artifically load up my test link with 100% load to keep it busy.
> everything appears to work, for about a minute. Then the link
> freezes for about a minute. No debugging info appears. Then the
> link starts running again for a minute and so on.
>
> I get the same result if I choose MPP encaps instead of MP and
> select "DBA Monitor=None" on the P50.
>
> Any ideas? Does anyone get good results with BACP and Ascends?
>
> Thanks!
>
> --
> -bmw | Double helix in the sky tonight
> | Throw out the hardware
> | Let's do it right -- Steely Dan; Aja
> -
> To unsubscribe, email 'majordomo@livingston.com' with
> 'unsubscribe portmaster-users' in the body of the message.
> Searchable list archive: <URL:http://www.livingston.com/Tech/archive/>

-
To unsubscribe, email 'majordomo@livingston.com' with
'unsubscribe portmaster-users' in the body of the message.
Searchable list archive: <URL:http://www.livingston.com/Tech/archive/>