retronetcall OCTOI status update

This commit is contained in:
Harald Welte 2022-11-19 10:52:46 +01:00
parent be39389a69
commit 07a878d23d
2 changed files with 4288 additions and 0 deletions

View File

@ -0,0 +1,101 @@
OCTOI (Osmocom Community TDM over IP) Status Update
===================================================
:author: Harald Welte <laforge@gnumonks.org>
:copyright: 2022 by Harald Welte (License: CC-BY-SA)
:backend: slidy
:max-width: 45em
== Overview
* Brief OCTOI Intro
* migration to co-located hub complete
* osmo-e1d status update
* new osmo-isdntap project
* OCTOI BERT problems
* Octoi Event PBX
* yate problems regarding call.answer
* resuming work on the BRI interface
== OCTOI in a nutshell
* community TDM network on top of the Internet
* to interconnect classic TDM and ISDN equipment in absence of public networks
* audience
** retrocomputing folks with interest in data communications
** hopefully at some point also museums
== migration to co-located hub
* noris.net provides free power/connectivity/colo for us (thanks!)
* lots of preparatory work to get GPS signal to "our" rack
** https://osmocom.org/issues/5542[OS#5542], https://osmocom.org/issues/5624[OS#5624], https://osmocom.org/issues/5626[OS#5626], https://osmocom.org/issues/5627[OS#5627]
* physical setup in mid-September
* C*NET ENUM routing change on October 1st
* Users all migrated (https://osmocom.org/issues/5690[OS#5690]
* https://osmocom.org/projects/octoi/wiki/Colocated_Hub
== software setup at co-located hub
* AVSt server
** DIVF VM: 2x Digium TE820 passed through via SRV-IO
*** runs `osmo-e1d` with dahdi-trunkdev for remote TDMoIP links
*** runs `yate` as simulated central switch
*** runs `osmo-isdntap` to record calls
** grafana LSC for https://stats.retronetworking.org/
** services LXC for future services (e.g. telnet services)
** openvpn LXC for remote access to event PBXs
** synchronet LXC to experiment with our own synchronet BBS
== `osmo-e1d` status update
* support for https://osmocom.org/projects/e1-t1-adapter/wiki/E1_tracer[e1-tracer hardware]
* osmo-1d now finally has a https://downloads.osmocom.org/docs/latest/osmoe1d-usermanual.pdf[user manual]
* rpm packaging (spec file thanks to @roox) for openSUSE Tumbleweed + CentOS8
== `osmo-isdntap` project
* we already can record raw E1 lines (e1-tracer, osmo-e1-recorder, ...)
** that was useful for getting a few minutes/hours of Abis traffic for OsmoBSC etc.
* for ISDN PRI use cases, we'd want
** pcap files of D-channel traffic
** raw captures of B-channel data
The b-channel recorded data should help us understand more of the B-channel protocols (X.75, T.70, G4 fax, V.110,
V.120, H.21, ...)
== `osmo-isdntap` project
* very simple hack at this point
* create RXMIRROR + TXMIRROR of DAHDI D-channel
* filter all Q.921/LAPD I frames
* simplistic decode of Q.391 message types
* detect whenever an exclusive B-channel is assigned (start recording)
* detect whenever a call is disconnected (stop recording)
== OCTOI event PBX
* A _modded_ Auerswald COMmander Basic 2 19"
** internal icE1usb for OCTOI connection
** internal beaglebone green (Linux SBC) for running `osmo-e1d`
Fully plug+play setup for having a few S0/BRI and analog PSTN ports at an event
One unit complete, three other units almost finished (waiting for icE1usb production)
== BERT problems
* https://osmocom.org/issues/5694
== yate problems regarding missing call.answer
* TBD
== resuming work on the BRI interface
* https://osmocom.org/issues/5417
== EOF
End of File

File diff suppressed because it is too large Load Diff