atomeara:
Looks like the roll out is coming mid to late May, about 160,000 lines listed so far.
Hoping that moves back to earlier in may like originally planned, vectoring was looking like a good birthday present from chorus :3
![]() ![]() ![]() |
|
atomeara:
Looks like the roll out is coming mid to late May, about 160,000 lines listed so far.
Hoping that moves back to earlier in may like originally planned, vectoring was looking like a good birthday present from chorus :3
#include <std_disclaimer>
Any comments made are personal opinion and do not reflect directly on the position my current or past employers may have.
atomeara:
Looks like the roll out is coming mid to late May, about 160,000 lines listed so far.
oh.
i wasn't in on friday and i filter those notifications.
crazy list... maybe dump into a fusiontable
#include <std_disclaimer>
Any comments made are personal opinion and do not reflect directly on the position my current or past employers may have.
Linux:cruxis:
Is chorus competing against enable in christchurch ? As fibre was rolled out in my area place Chorus changed the equipment in the cabinet unleashing my VDSL to over 108/44 which better than a equivalent fibre plan at the moment. But i am still getting fibre installed while its free anyways.
Can chorus make vdsl go even faster and undercut the faster fibre plans? How fast can VDSL actually go?
Fibre can't compete with copper when it comes to capacity and latency and I am sure many other things
Linux
Last time I copy and pasted a table on here, everyone complained about the table
atomeara:
Last time I copy and pasted a table on here, everyone complained about the table
Wrong quote?
Geekzone is pretty messy when it comes to tables, i blame @freitasm
#include <std_disclaimer>
Any comments made are personal opinion and do not reflect directly on the position my current or past employers may have.
Yes, tables are a pain. Copy to Notepad and copy from Notepad to the reply editor. Solved.
Please support Geekzone by subscribing, or using one of our referral links: Dosh referral: 00001283 | Sharesies | Goodsync | Mighty Ape | Backblaze
freitasm on Keybase | My technology disclosure
atomeara:
Yes wrong quote
fusiontable is an easy tableformat for google to transform into a map view too btw, just realized you thought i meant pasting a table here!
2114 dslams, that's a nightmare regardless of how you paste it into geekzone!
I'm pretty sure @BMarquis purposefully planned my cabinet towards the end!
on a more serious note, I'm keen to see vectoring on my personal lines, with G.INP i'll have reason to switch my adsl no interleavings line over to vdsl and enjoy the crosstalk being vectored (Very noticeable)
in terms of stats, from the last upgrade batch it's looking pretty impressive tbh.
some massive gains, not as many heavy losses - although RIP 50mbit upload lines!
Error rates are looking impressive! sadly resyncs overall are still higher than normal across board (although individual lines many saw substantual reductions so I'm yet to quite conclude that one - maybe single lines throwing off results with the storms?)
I'm still progressive through for an in-depth analysis - Could have Queried and left it at that, but outliers are sketchy in that instance.
Still seeing some dslams out there that have had their profiles switched under plans, but for whatever reason linecard isn't allowing it.
The only other oddity is the profile for HG630's which rather than 99mbit is set to 111882kbit - at which point atleast myself i say is pointless even defining as an exception.
#include <std_disclaimer>
Any comments made are personal opinion and do not reflect directly on the position my current or past employers may have.
https://sp.chorus.co.nz/product-update/vectoring-rollout-increase-speed-and-stability-vdsl
Chorus have also put out their own discussion on this now.
The graph they have presented, i would say lines up with my personal findings - unlike what was experienced with the previous PR. @ChorusNZ well done! :)
There are a few interesting bits in here, such as the discussion around minmum latency where G.INP is not working - This is important to those worried about latency and thinking 0 is better than a minimal 4.
The remaining lines that previously had an upstream speed greater than 30Mbps, saw an average 9.5Mbps decrease in upstream speed (24%). This group saw an average 14Mbps increase in downstream speed (15%).
Also commend chorus for making this stat black and white :)
#include <std_disclaimer>
Any comments made are personal opinion and do not reflect directly on the position my current or past employers may have.
hio77:
I'm pretty sure @BMarquis purposefully planned my cabinet towards the end!
Sorry, but I didn't put the schedule together, although that was an excellent piece of luck ;)
hio77:
The only other oddity is the profile for HG630's which rather than 99mbit is set to 111882kbit - at which point atleast myself i say is pointless even defining as an exception.
The sync cap for those devices isn't in place yet. It will be 99mbps due to 100mbps ethernet ports.
hio77:
There are a few interesting bits in here, such as the discussion around minmum latency where G.INP is not working - This is important to those worried about latency and thinking 0 is better than a minimal 4.
The best way to not be subjected to 4ms interleaving is to ensure your device supports G.INP. Fast re-transmissions at DSL layer is superior to iFEC interleaving - you only get re-transmissions (and the associated delay) at the point in time where it is actually needed. (during impulse noise events)
BMarquis:
Sorry, but I didn't put the schedule together, although that was an excellent piece of luck ;)
haha sure sure :P
BMarquis:
The sync cap for those devices isn't in place yet. It will be 99mbps due to 100mbps ethernet ports.
Considering my hit hitrate on data so far I'm still sure this limit is in there - possible just the identification was trialed here? (100% on all lines capable of more than 111.8mbit with hg630's)
BMarquis:
The best way to not be subjected to 4ms interleaving is to ensure your device supports G.INP. Fast re-transmissions at DSL layer is superior to iFEC interleaving - you only get re-transmissions (and the associated delay) at the point in time where it is actually needed. (during impulse noise events)
Fully agree, more current state discussion rather than when G.INP comes.
#include <std_disclaimer>
Any comments made are personal opinion and do not reflect directly on the position my current or past employers may have.
Pretty map, Done real fast so i haven't colour coded dates or anything.
When you see it mapped, you see the impressive scope on rollout - 10 days to hit all those dslams isn't bad!
This is purely based on the notified golive dslams.
#include <std_disclaimer>
Any comments made are personal opinion and do not reflect directly on the position my current or past employers may have.
Looks like there doing UFB 2+ towns but not UFB 2
hio77:
atomeara:
Looks like there doing UFB 2+ towns but not UFB 2
It's amazing what you can see when it's visual right?
yeah,
Interesting that they aren't doing Whangarei! Give it is other LFC
|
![]() ![]() ![]() |