Hi all,
JFYI: There's some progress on the nanoBTS multi-TRX configuration.
It seems it was my mistake. Setting the TRX0 unit ID to 1801/0/0 and the TRX1 unit ID to 1801/0/1 works using ipaccess-config.
If you then connect the TIB cable (TRX0 TIB out -> TRX1 TIB in) and reboot both nanoBTS, you will get a single A-bis OML connection from 1801/0/0 where you get software activate requests for both transceivers.
Somehow we're not handling them correctly yet from our abis_nm code, but that should be easy to solve. I'm working at it.
So Dieter and others: No need to worry about this at the moment, we're very close now and I think it is no problem for the 26C3.
Regards,
On Mon, Oct 19, 2009 at 08:48:48PM +0200, Harald Welte wrote:
Somehow we're not handling them correctly yet from our abis_nm code, but that should be easy to solve. I'm working at it.
I've now fixed a number of bugs, and the input/ipaccess.c code can now deal with multiple RSL connections with each their own stream identifier. There's also vty support to configure the stream identifiers.
The OML for both transceivers is coming up quite nicely, and the BTS also ACK the connection of the second RSL link.
However, I never actually see any packets from that second OML link, and the state of all the TRX1 baseband transceiver, radio carrier and channel objects still remains in "dependency"
I'll leave it at this for now. If somebody is able to provide an OML trace for a nanoBTS multi-TRX configuration, I would be extremely interested. I'm sure there's just some minor detail that I didn't know.
Regards, Harald
-
Harald Welte