Hi Neels,

On Mon, May 29, 2017 at 11:57:02PM +0200, Neels Hofmeyr wrote:

Just now I obtained a successful test run of the osmo-gsm-tester with these combinations:

OsmoNITB with sysmoBTS OsmoNITB with osmo-bts-trx (Ettus B210) OsmoMSC + OsmoBSC (AoverIP) with sysmoBTS OsmoMSC + OsmoBSC (AoverIP) with osmo-bts-trx (Ettus B210)

this is great news, and I hope we are getting stable, reproducible results for this. Might it make sense to simply let the tester run repeating and repeating the same test with the same software over and over again to see if it really is reproducible?

In any case, I think next to extending this to more BTS hardware options, there are the following next steps: * replicate the development setup on the 1U enclosed production setup * use the production setup for the official jenkins job * extend test coverage to go beyond the current very basic tests cases.

Would you agree?

I think a long time once did quite some brainstorming on the kind of testing that I'd love to see. Let me see if I can dig that out somewhere again. I think there's quite a bit that can be tested even without voice, purely on the signaling plane.