Quote from: npr

I believe if your routers stats shows the connection to be fastpath (interleave depth = 1) together with any  FEC errors then it's likely G.INP is active. If you think it's enabled on your line can you please post your router stats for us all to see. Some info here http://forum.kitz.co.uk/index.php?topic=14867.0

Quote from: Terranova667

with fibre the stats are gotten from the modem in which the BTO modems are locked down, unless you have managed to get hold of a BTO modem that has been unlocked or bought your own all in one you wont be able to see the stats yourself to see if G.INP is active. someone on the CRT could run a GEA test and see if it's enabled it would say Retransmission Low/High rather than Interleaving if it's enabled ( got that from the think broadband forums)  I wouldn't mind knowing that myself, i had my ping drop early this morning from 20-24ms down to it's usual 12-15ms could simply be DLM removing interleaving or something else i dont know

Quote from: WWWombat

I think this "FEC Errors with fastpath" is a red herring, and is not a sign of G.INP being activated. It is perfectly possible for the modems to make use of FEC correction without interleaving - they are two different processes, although their usage tends to be interlinked. There is no point in having interleaving without FEC, but there is a point in having FEC without interleaving. I posted some analysis on one case of this near the end of that Kitz thread linked above. The stats from the modem in that case showed every indication of FEC being turned on, and no indication of G.INP being turned on. And FEC without interleaving was proving to be very effective on that line. What we don't yet know is *why* the modems end up with FEC configured without interleaving. There is little published, and a lot of guesswork is involved in decoding what BTO are up to. As for G.INP: I have seen one case of modem statistics that made it clear G.INP was in use - but it wasn't an HG612. That modem showed statistics for G.INP-specific things (with labels that included the text "G.INP"). It also configured the line slightly differently, where "bearer 0" was the standard data channel subject to G.INP retries. However, the modem created an additional "bearer 1" that is used by G.INP to send the re-transmitted packets. Bearer 1 was configured with low speed, but very high FEC and interleaving protection. I think this is to make sure that only a single retry is needed.