...and having sunk as low as 7 Mbps last night this morning it has miraculously recovered to 64Mbps.

Tap3 Test 5 21.35 13/6/15 Result 7.01Mbps down - 2.07Mbps up
Tap3 Test 6 00.30 14/6/15 Result 22.29Mbps down - 0.98Mbps up
Ordinary Test 01.05 14/6/15 Result 55.31Mbps down - 11.29Mbps up
Ordinary Test 06.48 14/6/15 Result 60.86Mbps down - 13.01Mbps up
Ordinary Test 08.25 14/6/15 Result 60.73Mbps down - 13.45 Mbps up
Ordinary Test 11.30 14/6/15 Result 56.59Mbps down - 13.56Mbps up
Tap3 Test 7 13.05 14/6/15 Result 36.72Mbps down - 0.89 Mbps up
Ordinary Test 16.38 14/6/15 Result 57.53Mbps down - 14.24Mbps up
Tap 3 test 8 16.59 14/6/15 Result 13.96Mbps down - 2.24Mbps up
Tap 3 Test 9 18.06 14/6/15 Result 12.46Mbps down - 2.17Mbps up
Tap 3 Test 10 19.42 14/6/15 Result 9.32Mbps down - 1.94Mbps up
Tap 3 Test 11 22.15 14/6/15 Result 7.42Mbps down - 1.84Mbps up
Tap 3 Test 12 23.00 14/6/15 Result 10.01Mbps down - 2.12Mbps up
Ordinary Test 07.35 15/6/15 Result 64.66Mbps down - 14.21Mbps up.

I must be imagining the pattern I am seeing here, because... well I could swear it only seems to be an issue at times when many people are probably on line... and goes away when things would be expected to be quiet...

Back on the phone to Vivaciti today and see it I get hold of a supprt member who can CBA to push BT for the current provisioned SFBB capacity at the cabinet and what the utilisation is......

Edited by deleted (Mon 15-Jun-15 07:51:34)

In reply to a post by uno:

Ask the provider to get BT to detail the current provisioned SFBB capacity at the cabinet and what the utilisation is.

Matt

There are precisely zero cabinets on the entire Openreach real estate that are maxing out their backhaul according to most recent information, Matt.

Maybe so but this isn't always the case hence the reason suggested they check.

Although "maxing out their backhaul" and "current provisioned SFBB capacity" are very different things.

Matt

So the parent exchange of Pembroke, Haverfordwest, has had some capacity issues historically.

BT Wholesale also apparently maxed out their capacity to the Openreach GEA network on the exchange.

I'm unsure whether a new CableLink has been installed yet. This may be delayed pending hardware and software upgrade of BT Wholesale's Multiservice Edge Router at Haverfordwest in the early hours of Wednesday of this week.

This would explain performance issues.

These issues would potentially be affecting customers on:

Haverfordwest
Pembroke
Johnston
Milford Haven
Neyland
Manorbier
Clynderwen
Angle

In reply to a post by uno:

Maybe so but this isn't always the case hence the reason suggested they check.

Although "maxing out their backhaul" and "current provisioned SFBB capacity" are very different things.

Matt

Matt, the links from the cabinets do essentially nothing besides backhaul SFBB.

If you can find any cases where a cabinet has maxed out its backhaul to the handover point please do share - Openreach seem to be under the impression it hasn't happened, which seems reasonable given it'd require users to be pushing an average of 20Mb/s - more than 20 times the average peak usage for an SFBB customer.

I know this.

However, the bandwidth provisioned to a cabinet is often far less which is then raised in steps based on the utilisation.

To stress, I am not on about a cabinet maxing out the backhaul but hitting a provisioned limit that can be raised on usable bandwidth.

(Some background. Customer on FTTC, poor speeds at "peak times". Suspected as a SFBB capacity issue. Pretty BT graphs followed showing average utilisation and provisioned SFBB capacity. After many fights, they admitted it was hitting the upper limit and causing "congestion". Provisioned SFBB capacity was raised the next day resolving the issue).

Matt

In reply to a post by uno:

I know this.BT graphs followed showing average utilisation and provisioned SFBB capacity. After many fights, they admitted it was hitting the upper limit and causing "congestion".

LOL

That reminds me of the converse.

Many years ago Harold Wilson, then PM, went on TV saying his aim for the education system was that all pupils would achieve above average levels.

In reply to a post by uno:

I know this.

However, the bandwidth provisioned to a cabinet is often far less which is then raised in steps based on the utilisation.

To stress, I am not on about a cabinet maxing out the backhaul but hitting a provisioned limit that can be raised on usable bandwidth.

(Some background. Customer on FTTC, poor speeds at "peak times". Suspected as a SFBB capacity issue. Pretty BT graphs followed showing average utilisation and provisioned SFBB capacity. After many fights, they admitted it was hitting the upper limit and causing "congestion". Provisioned SFBB capacity was raised the next day resolving the issue).

Matt

You're talking about SVLANs, not cabinet backhauls. SVLANs go from BTWholesale's exchange edge routers to the rest of their network.

These are conveniently prefixed SFBB- in the reports.

The Openreach cabinet backhauls are point to point GigE links dedicated to taking traffic from a cabinet to a port on a switch. Not a lot of point in limiting uncontended links.

The BT Wholesale SVLANs are a different story; they share physical links out of the local MSE so are rate limited and those limits raised as required to make better use of the physical links.

Indeed.. but I never said I was talking about backhaul.

Matt

In reply to a post by uno:

Indeed.. but I never said I was talking about backhaul.

Matt

I have no idea what that has to do with what I've said.

All I said was that the SFBB capacity issue you were talking about related to BT Wholesale backhaul being split into SVLANs and it was on one of those where the issue you mentioned arose. Alongside that that cabinet backhaul is not rate limited.

Would you mind letting me know which bit I've gotten wrong? Thanks.