home *** CD-ROM | disk | FTP | other *** search
- Newsgroups: comp.dcom.isdn
- Path: sparky!uunet!mcrware!mcrware!curts
- From: curts@mcrware.UUCP (Curt Schwaderer)
- Subject: Re: Network Connections over ISDN
- Message-ID: <1992Aug14.171250.9284@microware.com>
- Sender: curts@mcrware (Curt Schwaderer)
- Nntp-Posting-Host: mcrware
- Organization: Microware Systems Corp., Des Moines, Iowa
- References: <1992Aug11.123306.145@indyvax.iupui.edu> <1992Aug12.085917.5543@ericsson.se>
- Date: Fri, 14 Aug 1992 17:12:50 GMT
- Lines: 46
-
- In article <1992Aug12.085917.5543@ericsson.se>, etxorst@eos.ericsson.se (Torsten Lif) writes:
- |> In article 145@indyvax.iupui.edu, imhw400@indyvax.iupui.edu () writes:
- |> >In article <1992Aug10.132038.26320@ericsson.se>, etxorst@eos.ericsson.se (Torsten Lif) writes:
- |> >[...]
- |> >> I'm more than just a little annoyed with how the marketing crowd make
- |> >> broad statements about the D-channel capacity being 16 kbit/s without
- |> >> ever mentioning how much of it is CONSTANTLY used by the traffic
- |> >> overhead. The D-channel's primary function is to provide a signalling
- |> >> path between the terminal(s) and CO. There exist signalling elements
- |> >> for user-user communication but they are (at the best) of secondary
- |> >> priority to the CO. Not that I dislike the idea of using the D-channel
- |> >> for user-user data transfer, far from it, but it should be made
- |> >> perfectly clear that this is not a high-speed/volume service.
- |> >
- |> >Okay, would you please help use newbies understand where this CONSTANT
- |> >signalling traffic comes from on a subscriber loop. What do the terminals
- |> >have to discuss with the CO when they aren't setting up or tearing down calls?
- |>
- |> When you put it that way - Good question!
-
- There are a couple of messaging protocols at layer 2 going on excluding sending
- call control packets for layer3.
-
- There are timers involved at layer 2. If no messages have been received in a
- time period (usually ~20 sec), Layer 2 will send what is called an "S" frame to
- the network (or vice versa) soliciting a response just to verify the link is
- still up.
-
- There are also TEI check and response messages that go on periodically that
- the CO initiates to verify and maintain all of the endpoints of a given
- ISDN line.
-
- All in all, the messages that layer 2 sends over the line are pretty sparce.
- The timer situation will only send messages on an idle line. The TEI check stuff
- is only initiated when the CO hasn't heard from a TEI in a long time. So
- layer 2 usually keeps the D channel pretty free.
-
- --
- +-----------------------------------------------------------------------+
- | Curt Schwaderer, WAN engineer, Microware Systems Corp |
- | Des Moines, Iowa 50325-7077 |
- +-----------------------+-----------------------------------------------+
- | curts@microware.com | My statements may not be UNDERSTOOD BY, let |
- | uunet!mcrware!curts | alone REFLECT Microware's viewpoints. |
- | (515)224-1929 x388 | |
- +-----------------------------------------------------------------------+
-