Ascend Archive
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

(ASCEND) Problems with T1<->PRI and Max4004


[ This message is long - you have been warned.  ]
[ Ticket opened with Ascend, but I've been wai- ]
[ ting way too long for a callback.             ]


Equipment:
	Max 4004 w/DM-12,BRI,Host/Dual cards, T1<->PRI installed (5.0Ap42)
	Meridian PBX (connected to Max via T1, D4/AMI)
	USR Total Control modem bank connected via PBX
	PRI circuit inbound from telco (ISDN signal, AT&T switch ESF/B8ZS)

Symptoms:

1) Calls to the USR Total Control modems (through the Ascend to the PBX)
   experience a pattern of delays.  This delay occurs if the T1<->PRI link
   is traversed, either inbound, outbound or both.  Pattern includes two
   delays of up to 14-16 seconds, as tested with ping(1); no packets are
   lost, but seem to be buffered.

2) Pattern is not present when calling DM-12 modems via the PRI.  Pattern
   is not present when dialing USR Total Control via DID trunks (bypassing
   the Ascend).  Pattern is not present when dialing USR via PBX (i.e. 
   dialing local 3-digit extension to reach modem bank)

3) We have two Pipelines making ISDN calls to the Max via the telco PRI.
   They do not show this behavior.

Attempted Resolutions:

1) Installed new Max 4004 per RMA.
2) Modified characteristics of Win95 DUN. (1.2b)
3) Upgraded to 5.0Ap42 (from 5.0Ap24)

Other problems:

1) Channels on the T1 between the Max 4004 and the PBX occasionally hang
   in "calling" state.  (i.e. status display is 'c')  The only way to clear
   these is to remove the T1 from the Max for a few seconds, and then re-
   connect it.  (These build up over time; I went 6 days without one, but
   have seen 6 channels hang in the last 24 hours.)

2) DM-12 modems (v.34) cannot maintain a connection in the absence of data.
   If I *immediately* start some sort of traffic generation (i.e. FTP ses-
   sion or 1000-ping barrage), the connection will stay up until the data
   ends; once the data stops, the connection is dropped within 10-30 seconds.
   (I have disabled all idle timers)


Supposition on delays:

The T1 connection is D4/AMI, which the telco informs me is 56K data and
8K signaling.  This is being converted to a 64K PRI data stream; the re-
verse is taking place on inbound data.  Is it possible that this signal-
ling change is dropping bits?  Is this translation particularly slow, cau-
sing (perhaps) these delays every 30 seconds or so?  The fact that these
delays do not occur when calling the DM-12 modems would seem to indicate
that the telco PRI is basically healthy.

Supposition on T1 channels "hanging":

I have no clue.

Supposition on DM-12 modem performance:

I'm at a loss on this one.  I've disabled every timeout I can find, and
they just won't hold a connection.  I read in the 6.0 Release Notes that
Ascend "has problems with IPX," but I see the same behavior on IP-only
PPP calls.  I've upgraded to the latest Win95 DUN (1.2b), also per the
6.0 Release Notes.  

Comments/suggestions welcome.  If you'll send your comments to me, I'll
summarize them back to the list.

--Wes

++ Ascend Users Mailing List ++
To unsubscribe:	send unsubscribe to ascend-users-request@bungi.com
To get FAQ'd:	<http://www.nealis.net/ascend/faq>
Follow-Ups: