This is merely a historical archive of years 2008-2021, before the migration to mailman3.
A maintained and still updated list archive can be found at https://lists.osmocom.org/hyperkitty/list/osmocom-event-orga@lists.osmocom.org/.
Harald Welte laforge at gnumonks.orgHi!
On Thu, Aug 29, 2019 at 04:26:51PM +0200, bibor wrote:
> A non representative survey lead to the date
> Monday, 2nd September at the c-base beach.
I'm not available then, sorry.
My feedback in terms of possible areas of improvement is:
* automatic saving coredumps and corresponding binaries/libraries is a must on the server
* there should be clear procedures about how to roll out patches during
the event, stopping the systemd service and running a custom-built
binary from a branch that's not rebased on what was running before is
not the best approach. It's not a desaster either, but we should know
better, IMHO
* LTE core network needs introspection (like our VTY). I'm currently
discussing with Sukchan what's the best approach here.
* SMS over LTE should have been possible (via SGs interface, implemented
by both OsmoMSC and nextepc-mmed) but I had no energy left to try.
Should be rather straight-forward in any future installation.
* next logical step in terms of 4G <-> 2G/3G integration would be a
combined GGSN and P-GW, i.e. something speaking both GTPv1 and GTPv2.
Not sure if we'll find any time to implement that soon (or try ergw),
but it would be the precondition to mobility (cell reselection or even
handover) between 4G and other technologies
* LTE core network shouldn't run on a laptop whose lid can be closed
(and as a result, suspended). I simply dind't have any non-laptop
machine available. Given that LTE has actual user data throughput
that's worth mentioning, we cannot use APU/Atom/Raspi for it.
Also some news that maybe haven't reached wider circulation yet:
1) sipgate (a German MVNO whom I know from some previous consulting
work) has indicated interest to have their SIM cards roam into our
event networks in the future. There are technical details to be
sorted out, but at least mobile data could be relatively easy with
the new GSUP/diameter translator.
2) One of the sysmocom customers (a small MNO) has offered to provide us
with diameter access to GRX, which would technically allow us to have
virtually any user roam to our network. The problem is that diameter
is signaling only, and we would basically disrupt voice and/or SMS
services for such users, which clearly is not a great idea. If we
want to do something like that, there would have to be an explicit
white-list to ensure only people who know the consequences will be on
the event network, and not uninformed bystanders.
I will stay in touch with those operators and keep this list posted
about plans. I don't forsee any significant changes to the existing
Osmocom network structure to enable any of the above. It's basically
a question of either using OsmoHLR to route GSUP messages based on IMSI
prefixes, or to have a dedicated GSUP router in between OsmoHLR and
OsmoMSC/SGSN. The router would then make sure that only 26242 ends up
in the OsmoHLR, while other prefixes (MCC/MNC) are treated by the
diameter translator and possibly routed to other operators.
Regards,
Harald
--
- Harald Welte <laforge at gnumonks.org> http://laforge.gnumonks.org/
============================================================================
"Privacy in residential applications is a desirable marketing option."
(ETSI EN 300 175-7 Ch. A6)