Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[doofus]

Hello all,

a bit of a weird one for you here!
On my brand new 330/50 FTTP line with AAISP my base latency is not stable, the BQM graph is is not a flat green section like it should be.

I am using the same router as I did on my previous FTTC line which did not exhibit this problem.
The BQM shows regular ramps up and down in base latency, and are also visible if you zoom in on the AAISP graphs too.

I know it is not service affecting, but it is definitely "not right".
I have swapped routers & rebooted the ONT, and it has not fixed it...

Anyone have any ideas, maybe the ONT is faulty?

[doofus]

There is another thread on the forum about a similar saw tooth BQM pattern:

https://forum.kitz.co.uk/index.php/topic,27103.15.html

But no solution is mentioned... 

[g3uiss]

Yours looks fine and similar to mine.

[jaydub]

Yours looks fine and similar to mine.

Mine shows the same pattern, originally on the move from IDNet FTTC to FTTP and is the same with AAISP, although the base ping time is slightly better now.

[g3uiss]

Is this causing any issues with your circuit. .?

[doofus]

No issues are being observed and the circuit seems to work fine.

I just find it strange that the last FTTC BQM is very different, despite using the same identical router:

I wonder why the shark fin / sawtooth pattern is showing on FTTP but not FTTC, is it a G-PON thing?

[XGS_Is_On]

Neither my Huawei or Adtran show that behaviour and not seen much of it so not a general FTTP thing.

[doofus]

I believe I am on a Nokia build, this was the overnight BQM:

[bogof]

I don't see it on my Nokia ONTs (I have both a 1G and a 2.5G single port ONT).
I was on AAISP previously, now on Unchained ISP, both with 1G services.
I'm using an OPNsense box as a router.

I do find that my base latency ping varies by around +/-1ms across router reboots - I assume this is owing to the connections getting to the ISP via different routes through BTW's network.  I've seen this both with AAISP and Unchained, who are both BTW based.

[doofus]

Hi all, I have opened a support ticket with AAISP, which a guy called Stuart has picked up.
A circuit self test shows:

Code: [Select]

BT Test E2E Access Test/DCN:Pass OR test pass. No fault found.
Down: (//) Up: (//)
Up Pass:GTC_FTTP_SERVICE_0000 Service is OK in the Openreach network. Please check the customer home environment set up.


[doofus]

So I ordered a regrade to 1000/115 and the sawtooth issue has now disappeared.
It seems the BQM wave/saw tooth effect was caused by the line rate cap when it was at 330/50.

The speed regrade happened just before 2am:

[bogof]

Might or might not be the rate cap.  A couple of other thoughts:
1) AAISP do shaping at their end I understand which you can customise the % of line rate they shove into your connection, I wonder if any of those settings would have changed things.
2) In moving from 330Mbps to 1000Mbps you also probably moved onto different hardware at their end, I understand the 330 and lower services are on FB6000, the 550 and above on FB9000.  I wonder if that might have had some effect.
Anyway, enjoy the upgrade!

[kitz]

There's also the possibility of a rounding issue that could amplify this.  We are talking extremely short periods of time and some lines do have minute jitters so that 15.4ms displays as 15ms and 15.5 as 16ms

[Alex Atkin UK]

Bottom line, unless you are experiencing real-world problems, its pointless chasing the perfect graph as its only intended to be a guide when you are actually having real-world problems.