In reply to a post by Ignitionnet:

I'll ask a few questions about Crawley and see what I can find out about why the first round of upgrades were inadequate and what the next steps are.

Questions answered - whole bunch of stuff going on which I'm sure you'll be informed about tomorrow. Capacity being delivered in 2 major ways, both of which need new kit.

Edited by deleted (Wed 04-Feb-15 15:52:37)

In reply to a post by mlmclaren:

3% and these figure come from where exactly? about 3% of the city I live in is actually getting steady speeds above 100Mb/s at all times of day, and I know of a big chunk of Birmingham suffering terribly to and has done for years, also Dudley's ongoing issues.... That's more than 3% right there.. It must be... and I know there's more areas also suffering

The issues affecting parts of Dudley and neighbouring areas are partly sorted now. They were delayed by the need to upgrade the environmental control in the head end but nodes are now being split onto the newly installed capacity alongside additional channels being brought online.

In reply to a post by mlmclaren:

my areas has had lots of new downstream channels added, my SH has had a whole new bunch of them but still issues, and investment over the next 3 years is too late... They should of been doing it continuously to prevent the issues customers are faced with.

Crawley had some nodes split. This is not the same as additional downstream channels though, that's a different upgrade and would show as something different than new channel IDs. Those don't actually mean much in the grand scheme.

Crawley has some quite odd network topology which makes it time consuming and expensive to upgrade it past the current state. It's being done but it does take time. Usage increased more rapidly than forecast and upgrades couldn't be done as quickly as would've been liked.

Edited by deleted (Wed 04-Feb-15 22:51:06)

I'd be interested to know how you run the automatic tests and how you manage to ensure no other use of the connection while the test is running.

can I ask you one question? (sorry two questions)

Why do isp's repeatedly under estimate utilisation increases? I keep hearing the same thing again and again, utilisation catching them out. Its as if they stick to the weird myth that if you increase end user speeds, they wont use more bandwidth, even tho it gets proven wrong time and time again.

Thanks for answering my question regarding what VM count as congestion tho, do you know if they have changed the thresholds since change of ownership, or do they still require extreme utilisation to qualify for a upgrade?

Edited by Chrysalis (Thu 05-Feb-15 22:41:19)

In reply to a post by Chrysalis:

can I ask you one question? (sorry two questions)

Why do isp's repeatedly under estimate utilisation increases? I keep hearing the same thing again and again, utilisation catching them out. Its as if they stick to the weird myth that if you increase end user speeds, they wont use more bandwidth, even tho it gets proven wrong time and time again.

Thanks for answering my question regarding what VM count as congestion tho, do you know if they have changed the thresholds since change of ownership, or do they still require extreme utilisation to qualify for a upgrade?

Sure.

1) It's not the end user speed increases that are the issue, it's the increasing usage of streaming services, alongside increased usage of digital media over physical media that's the issue. I can't go further into detail on the how / where / why than that.

2) VM adopting LGI planning guidelines? Quite the opposite if anything. Those are not the points at which network is considered congested but the points at which it qualifies for escalation for upgrade. Ideally this would never happen but obviously ideal and real are very different.

VM do use Samknows data as well as customer reports and their own KPI utilisation reports to assist with capacity planning. Most of the time the reason there are issues is because of delays in installing new capacity. When a port is escalated there usually were actually already plans to upgrade it, along with many others, as part of a programme that has been delayed.

As mentioned, in the case of Dudley this was due to the environmental control needing upgrade to allow new install of kit. In some other cases it's been things like the hubsite literally having no room left and VM having to build an extension to house more equipment, or rearrange things while trying to minimise disruption.

Things like power needing upgrading, far from trivial, or even vendors having problems with production of equipment and hence it being unavailable for a time can also stall things.

Capacity planning know the status of every port on the network as a general routine. They obviously don't treat the upgrade of every port as an individual project, they will have many projects running consecutively, however a single port can certainly be escalated and have a fault raised against it.

Things have come a very, very long way.

No comfort, of course, to people affected by issues sadly.

In reply to a post by Chrysalis:

Why do isp's repeatedly under estimate utilisation increases?

Because they want to save money, despite getting more from as many customers as they can cram onto their links.

Apart from that it's poor planning. ISPs will make out they have dedicated teams for those kind of faults/upgrades but those teams are not very effective.

Not quite sure on the other question. I do know though that there are some areas around the country, especially Bristol, where congestion is huge (Or was a year or two ago). I've never seen them be very fast when upgrading the capacity of a congested area though.