On 09 May 2016, at 11:49, Neels Hofmeyr
<nhofmeyr(a)sysmocom.de> wrote:
On Sun, May 08, 2016 at 05:13:51PM +0200, Holger Freyther wrote:
Have you considered removing the
"timeout" from the lower GSM paging code? The interface would be start, cancel
and only success would be signaled? Both the MSC paging and BSC paging timeout could be
fed from the same timeout value?
So the MSC tells the BSC to start and stop paging, such that the T3113 is in
the MSC? I'd expect the network design to be intended otherwise, to reduce the
amount of work towards the CN (handle as many details as possible in BSC).
No the MSC just commands to page. But it has an internal timer when it gives up and even
if there is a paging response will not handle it anymore.
And we want a way that works both with 2G and 3G. Can
we tell an RNC to start
and cancel paging? Not that I know of ...?
Neither. So it makes sense to have the timeout in the lower layers and one in the
MSC/subscriber layer. Only the subscriber layers' timeout should trigger the callback.
That is my opinion for the long run architecture.
Another thing: I guess that osmo-bsc should not change
behavior with respect to
3rd party MSCs, right?
right.
top-up? refill? resume paging?
/me favors 'resume_paging'
but whether we should spend time changing that is on another page.
true.