Terrible performance Leicester (LE3) - Page 7

Ignitionnet · 29-11-2010, 16:59

Quote:

Originally Posted by Chrysalis

I meant not upgrading to reduce to that level but rather that original sales werent enough to get higher usage.

You misunderstand how the network works. x customers are shared between y line cards. Where x is lower y is also lower to keep the number of customers per line card reasonable. There is no set formula for how many premises each line card covers, this varies depending on Virgin's needs.

They don't just throw in a line card for each 250 enabled homes and hope for good take up, the networks were originally thousands of homes passed per line card, even a 10% take up rate would have been ample to get 400 customers on a card.

---------- Post added at 15:59 ---------- Previous post was at 15:57 ----------

Quote:

Originally Posted by Chrysalis

why is legacy run at a different contention to the overlay network? thats an interesting one.

to me contention ratio is always relevent, its the ultimate measure of budgeted bandwidth. You said it yourself, the higher contention is only barely achievable due to STM and shaping been used to cut costs.

To answer the first point that's easy, because the overlay network has far more bandwidth available. It can push 200Mbps to each fibre optic node while the maximum that the VXRs can push is 76Mbps.

As I mentioned contention ratio isn't relevant - that a ratio far higher than would be feasible is doable cheapens the value of that metric.

If you look at ISPs with strict caps and pay per GB overages they run on huge contention ratios because they can. Visible contention is what it's all about and is the only game in town. Whether the contention is made less visible due to a lower ratio, shaping, STM, whatever, that's the only valid metric.

Chrysalis · 30-11-2010, 08:41

Thanks for the info so it seems legacy loses out quite a bit from overlay, in both a more fragile capacity size and over double the contention. I will defenitly not go back to legacy now if ever given the choice.

ignition interesting now that VM have released dates, leics has the latest possible aug 2011, I guess the build isnt as new and clean as you first thought? This makes what you said as unreasonable, if these upgrades do fix the area, another 9 months is far too long to wait.

I do fail to see however how STM and shaping reduce visible contention, they actually introduce it.

Chrysalis · 10-12-2010, 23:07

getting beyond a joke now.

service is seriously broken, I can load this site much faster on my adsl (which is fully utilised downloading) than on my idle so called 20mbit connection.

my up to 8mbit adsl service faster than my up to 20mbit VM service.

Chrysalis · 11-12-2010, 14:55

Philce I have given VM 6 weeks notice (end of jan 2011), I have it free until then.

I will be getting a new adsl isp to replace ukonline when it closes, and have 3G as backup.

VM itself is way too variable and generally is slower for everything except bulk downloading (which itself is only faster mornings).

I have overlapped this and when I will be on my new adsl isp incase I have problems on the adsl and change my mind again.

Brief history of this current stint with VM (ntl) 20mbit service.

feb-may 2010 - high latency/jitter in day, low at night. speeds in day almost 0, often stalled, timeouts, streaming broken etc. unfit for purpose. unable to hit max speed even at 4am.
june-august 2010 - moved to overlay and things improved, speeds generally high, streaming etc. worked fine. 18mbit+ but still had high latency outside of dusk hours.
sept-mid oct 2010 - service was decent, jitter varied but was generally at least ok. speeds stable.
mid oct - now - service declined very fast and on top of this VM introduced traffic shaping which has brought extra problems, had mtu issues, browsing issues, streaming issues on way back, very inconsistent speeds even off peak, on peak starting to get very slow again approaching the 10% mark along with very slow browsing. jitter getting to the point where latency sensitive apps become unuseable. this time the slow speed is alongside VM shaping down p2p, nntp which should be an embarrasement to them.

proposed fix date of april 2011 by VM, likely to be delayed at least a month, I expect service would be adequate from may to august where it will then blow up on the speed upgrades. too much for me.

Chrysalis · 12-12-2010, 22:41

tonight is just completely broken.

not using it for generic use now but I seen my ping graph was crazy high all day so tried to use it.

web browsing missing images galore, connection reset by peer error's, same error in ftp when trying to download.

I managed on 3rd attempt to finish a speedtest and it took 10 minutes.

Chrysalis · 13-12-2010, 03:32

here is fastest (channel 54) at 2.30am

the other 2 channels 55 and 56 still sub 2mbit.

Chrysalis · 14-12-2010, 12:42

11.40am on weekday.

Chrysalis · 14-12-2010, 17:55

sloooowing down.

if I am up and can be bothered will do another 8-10pm.

qasdfdsaq · 14-12-2010, 21:05

Umm sorry if you've already answered in this long thread but, why haven't you left VM already?

Chrysalis · 14-12-2010, 21:11

I tried to give VM a chance due to having a poor line (which means a poor adsl service). I have now actually cancelled tho and VM will be gone end of jan. I am chaning adsl isp during jan so want VM up during that time incase of long adsl downtime.

I can still report here tho in the mean time

Me just leaving feels wrong tho, in that scenario VM get away scot free for what they doing other than losing my subs.

qasdfdsaq · 14-12-2010, 21:34

Ah, good to know. I was in the same situation a little over a year ago, and didn't regret leaving VM one bit.

But as is always in the corporate world, it's far easier for them to screw us over than the other way round. Same with a billing error - you can't stop them taking it out of your account and it'll take months to get them to refund you.

philce · 15-12-2010, 00:00

Id agree with you, but my speedtests are always near the 10meg I should have.

I have loads of stalled web pages, missing images and general slowdowns. I cant say anything to VM though as they will look at the utilisation and say its all OK. I may give the CEO office a call though as things are definitely getting worse.

My other option is a equally dodgy ADSL line (sub 4Mb if i'm lucky) that took me 2 years of BT faults visits to get me up to this speed from 2.5Mb, god only knows what has happened to the line in the mean time.

Latest Ping test

TBB ping test (dosent look that bad)

qasdfdsaq · 15-12-2010, 00:33

That's interesting because I seem to have observed VM's network/equipment appear to have two operating "modes" - one with low latency, high speeds, and considerable packet loss just as you describe, and the other one with moderate latency, moderate speeds, and zero packet loss.

What I observed yesterday while I think they were doing some capacity upgrade work, was that at one point I reconnected to a different upstream channel to usual. At this point, I got slightly higher speeds than usual, and also slightly lower pings. But packet loss shot up from zero to about 10%, stalled web pages, etc. as you describe.

Then about an hour or so later, my ping increased again, and packet loss dropped to near zero in an instant - while still on the same channel. Clearly, something changed, whether deliberately or not, and not gradually either. And it wasn't due to congestion or high usage.

If you look at the that graph, between about 9:30am and 10:45 packet loss was consistently high, but at 10:45 my ping increases and packet loss almost disappears (save for the further two reboots/loss of service at 11:00/11:20.

Incidentally during the periods of high packet loss, my ping would stay low. Normally when I saturate your upstream doing a big upload, my ping would increase to the 500-1000ms range but still not much loss. Yet in this period my ping would stay at ~30ms even while saturating the upstream, so it looks like something to do with VM's buffering/scheduling parameters.

I can't really explain it, but put simply, something VM did - some setting they may have been tweaking somewhere - managed to cause 10-15% packet loss all of a sudden, in some way that was completely unrelated to load, oversubscription, or line/signal quality. Maybe a QoS or timing parameter, maybe something else. But still, your problems might not be completely capacity related.

Chrysalis · 15-12-2010, 01:07

yes I very rarely had packetloss but the latency/jitter is through the roof and speeds very inconsistent. But in the last few days I am starting to see low packetloss.

philce's graph looks very good with the latency now, as I said on the VM forums his port improved at the same time mine got worse and I wouldnt be at all surprised if it was a reseg that moved a busy node from his port to mine.

trying to run speedtest now but keeps timing out, if it finioshes will post image.

---------- Post added at 23:56 ---------- Previous post was at 23:43 ----------

ok here

running these was fun, lots of retries and very slow.

---------- Post added 15-12-2010 at 00:07 ---------- Previous post was 14-12-2010 at 23:56 ----------

post midnight it becomes useable.

Chrysalis · 15-12-2010, 10:39

jitter is amazing today but speed maxing out flatline at 10mbit.