UK connections very poor - Anyone else in UK suffering?

[Blagard]

I ran pingplotter (206.16.60.39 and then 206.16.60.41) just before posting showing the MA at 400+ 

I then went online to AH to see what was in the clipboard for arena "ping" - It was just the same on all except the two special arenas that looked normal.

For a few weeks now I have noticed that the problem is consistent through late evening in the UK and only drops down to sub 200 when most in the UK should be in bed! - I appreciate this is not an AH made problem and wonder how many others in the UK are suffering likewise?

The strange issue of the two Special arena's showing a normal (for the UK) ping of around 150 - 170 is something I don't understand.

I ran ping plotter again after typing the above to give an example, it was better but still bad

[RTHolmes]

fine here:

[Blagard]

Cheers RTHolmes, I see you don't get hopped through Tinet

I really don't know how this part of the internet works, but some-one is paying for the service somewhere. I have had a polite whine at VirginMedia (My ISP) in case they have to pay Tinet! It's 2am in the morning right now and I am still getting the same lag as I posted earlier.

I really don't fancy trying to fly with that sort of lag as it will just wind me up even more.

[Noir]

check this thread out

http://bbs.hitechcreations.com/smf/index.php/topic,310034.0.html

[Bruv119]

whatever that router on 37.37 is where all virgin media connections get routed through on the way to HTC,

i've been getting round trip times of 400-600 between 5pm-11pm GMT (it is usually 140-160 during the day) this past week but playability in game the DA/MA has been much more stable  .

Micky has been in communication with tinet and the latest is that they are pointing the finger at htc.   I however believe it is Virgin's traffic management system and or network during peak hours.   Their literature says they will not throttle you depending on your usage but I believe we are being bummed with the content sharers.  Although the hop in question is stateside other isp's seem to go through a similar block of IP's without the same sort of increase in trip time.  So unless there is some way we can get this tinet to admit that it is their problem we will have to talk to virgin.  (good luck with that and let us know how you get on   Ideally we need to be routed away from 192.205.37.37

I will also say that this has only started happening within the last few months,  so something somewhere has been tinkered with and results will vary.  I've considered getting a second ADSL connection but the symptoms seem to come and go.

[RTHolmes]

I had this problem a year or 2 back with Demon, they were using some Tata owned company for peering and the atlantic hop was always getting conjested then slowing down, then losing packets. even when the hop across the pond was ok there were delays in NY between Tata and AT&T. plus Demon's internal routing was becoming rubbish - too many hops and packet loss. after years of reliable connection it suddenly became pretty much unusable.

thats why I switched to Entanet (via one of their resellers Aquiss) because they actually seem to care about getting their peering right (enough capacity) and have the techies around to run their network properly, although that does cost more. last night's ping was pretty high for me, its usually ~120 with 14 hops.

We have 2x ADSL2+ lines at work for bandwidth sharing, one with Demon and one with Aquiss for redundancy:

Demon averages 4.5Mbit rx and 480Kbit tx
Aquiss averages 8.5Mbit rx and 950Kbit tx

... using the same modems, to the same exchange, in the same multicore cable ...

bottom line is not all ISPs are equal.

[Blagard]

whatever that router on 37.37 is where all virgin media connections get routed through on the way to HTC,

So unless there is some way we can get this tinet to admit that it is their problem we will have to talk to virgin.  (good luck with that and let us know how you get on  Ideally we need to be routed away from 192.205.37.37

This is what Tinet said when I sent them a pingplot the graph was more uptodate


"Looking at the graph you attached it results that the rtt within our
backbone is working within normal expectations (hop 3 is reached in 97ms).
We have escalated this issue to our peering department to have it
further checked."

I have absolutely no idea if they have any control on it going to the "37.37 "

[RTHolmes]

hop 3 is reached in 97ms

are they looking at the same pingplot?

[Blagard]

are they looking at the same pingplot?

  Yeah I'm sure they are! - It's just that they are looking at the hops they (tinet) are involved with. You can see the 97 one on the plot.

[RTHolmes]

89.149.187.109 is in germany, av ping to that router is 111 - from the UK thats dreadful (although apparently "within normal expectations")

[Blagard]

I would not trust those database's for information - the next hop is listed as in Middle of the USA. Wichita

[Raphael]

I am suffering here in brazil 
i have always had ping at 200 wich is good and not laggy at all, but ever since the last update my ping got to 400 + and also a lot of discos, yesterday i was able to play at 2 am when i saw for the first time in these weeks a tasty 200 ping, but the whole day it is always more then 400 ): impossible to play it.

[mechanic]

Mine has gone to pot. For years I have been lag free and steady ping but recently I've noticed my hitsprites do no damage. during the scenario I was shooting at bombers that looked fine to me then when I watched the film, some of them were on fire already, despite not being on fire when I was playing. I also had a plethora of 'Packet Ignored' text lines come up in any film form the last few months.

[Bruv119]

This is what Tinet said when I sent them a pingplot the graph was more uptodate

"Looking at the graph you attached it results that the rtt within our
backbone is working within normal expectations (hop 3 is reached in 97ms).
We have escalated this issue to our peering department to have it
further checked."

I have absolutely no idea if they have any control on it going to the "37.37 "

will be interesting to find out.  I based my assumption on the fact everyone with a Virgin connection goes through 37.37 no matter what.  tinet being virgin's peer in the states.   I've googled the IP address and with a little digging it causes lots of people a very similar lag related problem.  Either their peering cannot cope with the level of demand during these hours or they need to route it differently.   

Please update us again if they reply with anything useful. 

[Blagard]

Out of interest I decided to ping some West Coast Users to see what happens (Obviously the location of the Servers may be different)

The results were that the majority were OK (including sandiego.com, sandiego.edu) and they did not go via tinet or the 192.205.37.37 node - Basically good enough for AH (Maybe)!

However I did manage to get one, "sandiego.org" that took a  near identical route at the beginning and went through tinet and 192.205.37.37 with exactly the same problem. Which at the time of writing is even worse at 500!

So it seems to me that the destination has a significant impact on the route the data takes from the beginning. It begs the question, what is it about the HTC servers that influences the route from the UK. The next question is why do other UK ISP's route the traffic so differently without the problems VirginMedia seem to present. I did read about peering, and it does seem to be a minefield and way over the top of my head - Saterlite Alt!

One for Skuzzy - Is there anything we can say or do to address the problem with our ISP (Besides changing ISP)?