Progress so far
Using a 7800N and grabbing the firmware 1.06d from the Aussie billion site and upgrading.
It's all working well now:
Issue i had was when specifying the DNS servers it failed to allocate an IPv6 address, whether this is a PN issue or OpenDNS I don't know.
It's all working well now:
Issue i had was when specifying the DNS servers it failed to allocate an IPv6 address, whether this is a PN issue or OpenDNS I don't know.
I ran that test, large discrepancy between 4 and 6 speeds for me, though I recall reading something about IPV6 and Cisco causing issues, I'll try to dig it up.
/edit - A quote from Andrews and Arnold...
[quote=http://aaisp.net.uk/kb-broadband-ipv6-tech.html]Cisco notes
For at least one customer running native IPv6 on a Cisco 877 router, they had a huge disparity between IPv4 and IPv6 speeds. IPv6 file transfers were about 10 times slower than IPv4 transfers!
It turned out to be a Cisco IOS bug; once the IPv6 stateful firewall (inspect) had been disabled, IPv6 ran at full speed.
It is Cisco bug CSCtb10776 (Inspection drops NIC-segmented packets when WScale is on). It can also be worked around by disabling TCP window scaling on the affected hosts, but that can have performance implications of its own. It has been fixed in IOS versions:
15.1(0.18)T
12.4(25b)M0.13
15.0(1)M1.2
I'm using 15.1(3)T so it shouldn't be that affecting me. Running another test and the speeds changed, now IPV6 is faster than IPV4.. Third test about equal, so something I should keep an eye on?
/edit - A quote from Andrews and Arnold...
[quote=http://aaisp.net.uk/kb-broadband-ipv6-tech.html]Cisco notes
For at least one customer running native IPv6 on a Cisco 877 router, they had a huge disparity between IPv4 and IPv6 speeds. IPv6 file transfers were about 10 times slower than IPv4 transfers!
It turned out to be a Cisco IOS bug; once the IPv6 stateful firewall (inspect) had been disabled, IPv6 ran at full speed.
It is Cisco bug CSCtb10776 (Inspection drops NIC-segmented packets when WScale is on). It can also be worked around by disabling TCP window scaling on the affected hosts, but that can have performance implications of its own. It has been fixed in IOS versions:
15.1(0.18)T
12.4(25b)M0.13
15.0(1)M1.2
I'm using 15.1(3)T so it shouldn't be that affecting me. Running another test and the speeds changed, now IPV6 is faster than IPV4.. Third test about equal, so something I should keep an eye on?
My latest:
Maybe it's an issue on my side, I know Kaspersky tends to mess up the PlusNet speed test although usually it's by jumping the speed up 10 fold.
Maybe it's an issue on my side, I know Kaspersky tends to mess up the PlusNet speed test although usually it's by jumping the speed up 10 fold.
Final test for the night:
100% the same
100% the same
Up and running with my 7800N, scored 10/10 for both IPv4 & IPv6 readiness at test-ip6.com so I think I'm okay.
Read 'em and weep...
