Hi Anoop,
On Sat, Feb 18, 2017 at 12:03:37AM +0000, Anoop Singh Tomar wrote:
Posting question related to capability to implement user plane (GTP tunnel) for S1-U and Iu-PS interface between S-GW & eNB and SGSN & RNC respectively via osmocom kernel GTP-U module.
The current kernel GTP-U module has been designed explicitly for the use case of a GGSN/P-GW. This was what the original requirement was, and where Pablo and I received some funding from a customer to implement it.
The requirements for a SGSN/S-GW are completely different, and I don't think the current module is any good match for this. Please see the discussion starting from http://marc.info/?l=linux-netdev&m=148611438811696&w=2 for the rationale about this, specifically my post at http://marc.info/?l=linux-netdev&m=148638986910316&w=2
If you think otherwise, please let me know. I'd be happy to undeerstand why.
At present it seems if TE-ID is not valid packet is dropped in kernel GTP-U module.
This is a bug/mis-feature that has been fixed already and the patch is currently in net-next, waiting to be merged into the next linux kernel merge window: https://git.kernel.org/cgit/linux/kernel/git/davem/net-next.git/commit/drive...
Question - Is there any plan to support above in near-future in osmocom kernel GTP-U module or larger question would be to plan to enable usage of kernel GTP-U module in SGSN and SGW towards RAN ?
There are no such plans, at least not from Pablo and my side. I don't know about Andreas, but I think he is also primarily focussing on various (more complex) P-GW use cases, rather than S-GW. The module was developed to fulfill certain requirements under a customer contract. Neither Pablo or I currently have any customers who have above requirement, so there's no related agenda.
I would love to work on the kernel GTP code, but I cannot afford to do so in my limited spare time (which is already occupied by way too many projects getting too little attention), sorry.
In general, I don't think the current GTP "tunnel endpoint" module fits the S-GW/SGSN use case at all (see my linked post above). The S-GW/SGSN is not a node visible in Layer3 of the user-IP payload, it must not terminate the user IP tunnel. A completely different architecture should be used for the SGSN/S-GW use case, where it is all about forwarding encapsulated packets between MME and P-GW without decapsulating the inner (user) IP layer.
Regards, Harald