On Tue, Oct 29, 2013 at 11:01:24PM +0100, Sipos Csaba wrote:
You mean the local end release being broken in the BTS? How can I find this out? I see your point, that the release command has been sent, but we never got any ACK or something like that back. But how is it possible that it works for SAPI=0 and it does not for SAPI=3 ? These units were used by big organizations to improve the indoor coverage, I really don't think that the BTS or its firmware is wrong by design.
Maybe OpenBSC destroys SAPI=0 too fast, and there is not enough time for this:
Check rsl_release_sapis_from and the caller in _lchan_handle_release. First everything from 1 t X is released and the SACCH deactivated.
Play with that, or attempt to get a trace of a Nokia BTS with a Nokia BSC and look at there release procedure. E.g. if order or release requests kinds are different.