In reply to a post by Andrue:

In reply to a post by Alucidnation:

Thanks but i don't understand what 'block' actually means?

It means 'multiple' as in 'A block of flats'. With IPv6 you aren't assigned a single address. You are assigned a contiguous range of addresses. IDNet issues what's called a /48 (I think most ISPs do). What that means in practical terms is that we get assigned a range of addresses that all have the same values in the highest 48 bits. And yes, that does indeed mean that we get assigned a large number of address - 2^80 in fact.

i know with IPv4, you just connected and away you go, but from what i understand, ipv6 isn't as straight forward as that, and wondered why there was a limit to how many IPv6 addresses you could have.

Oh it's straightforward for most people. It's only if you need to expose servers to the outside world that things can get complicated. For the average home user it's just as 'plug in and go' as IPv4. Possibly more so thanks to stateless auto-configuration.

Brilliant, thanks for the detailed reply Andrue.

Just to add, with IPv6 it no longer makes sense to think of how many IP addresses you receive, because it's more than you could use (even in a singe /64 = one network).

Rather you receive a prefix allocation (or delegation) which determines how many networks each with globally valid and unique address ranges you could run (behind your router).

A /64 is one network
A /56 is up to 256 networks
A /48 is up to 65536 networks

That's in addition to the /64 (or sometimes a /127 or /128) used on the link between your router to the next upstream gateway.

The prefixes are what matter because they define both what subnets are yours and the routing information that can be summarised (aggregated) into bigger contiguous blocks.

In a way it's like going back to a IPv4 time when all organisations were implicitly assigned in /8s (one size fits all) as I don't think they ever expected to have even 100 organisations participate in an experiment in tech research, except this time we don't run out after issuing 220 prefixes.

A belated thanks for this.

It appears i still have some learning to do.

Ok, all up and running with IDnet, although for some reason, when the migration was carried out, DLM has applied interleaving on my line, knocking nearly 10Mbs off my sync rate, with an attainable of over 70.

Not a good start really.

The guys did say they would request a DLM reset but said it may not improve things.

Not sure why, as i was getting around 68 with my previous provider.

It is normal these days for DLM to be reset on a migration, and interleaving is applied by default. I would expect IDNet support to have known that.

Assuming you previously had G.INP active, it should return in 3-4 days without any intervention. Don’t do lots of re-syncs!

I wouldn't worry too much as same happened to me when I moved across to IDNet and I was starting to wonder about the decision.

Within 48 hours G.INP kicked in on the download and I was synced at 72 rather than the 67 with my previous ISP. Since then SNR has gradually been ratcheted down and my router now reports sync at full 80 download with an SNR of 2. The BTW site reports a max download achievable of about 77 to 78 and a profile of 74. I'm actually achieving an average of 70 down and 17 up which is about 8Mbps better on the download than the previous supplier..

Thanks guys.

I’m on an eci cabinet so I don’t think g.inp applies to those?

You’re probably right, but the interleaving could well be removed .

Ok thanks.

They said it can take up to 72 hrs to be done so I’ll leave it a week and see what happens.

IDNet moved me to BTw backhaul in mid-December and I was interleaved on downstream for exactly 48 hours. I'm also on an ECI cabinet.