Terrible performance Leicester (LE3)

[Chrysalis]

philce about the pm's they didnt really help me as such as I didnt pm any staff on here, I was asking around for the CEO number which noone on here had the current one, but eventually got it of a helpful guy on VM's own forums which I then passed on to yourself to help you out as well. I got moved back by ringing them up again after I got the number.

[philce]

OK, to be honest it wasnt as bad as I was expecting.

I will call them again tomorrow.

[Chrysalis]

your legacy graph dont look as bad as mine was and thats probably because your port isnt as bad.

seems to be some light congestion here today on downstream, max about 3.6 in dumeter vs 3.8meg and 28mbit on speedtests. so not 30mbit available across all 4 downstreams. Explains why 20mbit was so flaky without bonding.

[qasdfdsaq]

It's interesting how big a difference the channel bonding makes, though not unexpected either. Fitting 200mb of bandwidth demand into 50mb was always going to have worse contention characteristics than fitting 800mb of demand into 200mb (actual numbers will be higher).

All comes down to statistics really.

But anyhow I'm glad things are a lot better for you now, and it does look like congestion in your area is actually in the same ballpark as mine; I was getting the same speeds as you while previously on 10mb, and now on 4x bonded channels ~25-35mbps is what it'll hit most days during peak times.

[Chrysalis]

What I found odd about this is every speedtest is exactly the same result, ftp or speedtest.net, congestion will normally show some variation. But its the same on other pc so it is probably congestion.

yeah even with this slow down as long as it doesnt start declining again its a lot better than without bonding. If it falls below 27mbit I will then push the CEO office on it again so they keep been pushed and not allowed to think things are ok (27mbit been the magic 90% expected performance). Although as far as I know some kind of upgrade is still scheduled in march.

[Chrysalis]

speeds back up to 30mbit now.

[qasdfdsaq]

Sorry, deleted my last post as it was intended for the galway thread, I posted in the wrong place.

[Chrysalis]

yeah it made me post some stuff in wrong place also, editing now.

[qasdfdsaq]

Originally Posted by Chrysalis

speeds back up to 30mbit now.

Mine's at a fairly resonably 46.5 today:



Funny how our two connections seem to be behaving in exactly the same pattern despite being several hundred miles apart

[Peter_]

Originally Posted by qasdfdsaq

Funny how our two connections seem to be behaving in exactly the same pattern despite being several hundred miles apart

Purely coincidental nothing more.

[qasdfdsaq]

Actually it's not coincidence, there's a reason and a cause for everything, even if you're not aware of it

[Chrysalis]

I am seeing affects on browsing now and from what I can see all my traffic going through VM's telehouse link has packetloss.

0.28meg down on london speedtest and 29meg down on milton keynes. 30meg on ftp in europe.

---------- Post added at 17:24 ---------- Previous post was at 17:17 ----------

seems all back to normal now except speeds down to 28 again and speedio still flaky. I think speedio messed up ddwrt and made it affect other locations until its congestion cache expired in the routing table.

[philce]

Cant seem to stay connected for more than 20 minutes at the moment, have to reboot modem to get back online.
Been happening all weekend.
Is legacy this bad?

[Chrysalis]

I never had reboots on legacy (apart from when I signed up) but it had so many stalls of throughput it was in affect a dropout.

[philce]

yes same thing, needs reboot to get back.

---------- Post added at 21:50 ---------- Previous post was at 20:56 ----------

Looks like the reboots has just killed my router!

Just stuck with the WPS button lit all the time. Wont respond to the reset button at all.

This just gets better and better!!!

Chrysalis can I borrow your old router till I get a new one???