375 replies to this topic

[Clit Beastwood]

Bhael Fire, on 12 December 2014 - 03:17 PM, said:

So...are users of TWC pretty much boned? Is there something that can done about this? It's very frustrating.

Dunno about you - I tried to switch ISP's but they acquired the only local option, so I'm stuck. I get better hitreg when tethering through my cellphone. 320mbits down and crazy packet loss.. I'd settle for 50mbits down and good hitreg.

[Zordicron]

Jon,

Did you get a chance to look at my Frontier Com. issues? Am I on the bottom of your list of to do's because i am the only unlucky SOB that has to play through Frontier on the planet?

I posted a trace route on that other thread you had going.

I am pretty sure I get throttled badly during prime time. My ping hovers around 400. During the wee morning hours, like 1 am and on, I have seen ping as low as 60 on say, Sunday at 1 am when my area is almost entirely asleep.

I do not know if there is anything to be done with this, Frontier has a monopoly, I am too far out of town to get cable(which is actually good in my little town) and I have heard nothing but horror and travesty from people that tried cell connections in my area.

[Syphr]

Fired up to play tonight and could not connect. Checked and sure enough, dropped packets trying to go out a TWC internet Connection. Not to mention the ping times were in the 150 range. Not bad, but when over half of the packets are being lost, that is an issue.

Now, fired up a VPN and was able to connect and even got pings in the 75ms range instead of 125-200ms.

The issue isn't with TWC. it is the route that TWC traffic takes going through ovh.com. One of the routers they have is so over loaded it cannot deal with larger amounts of traffic.

If MWO really wants to fix it, they need to get their hosting to change the routing to something better. Or at least find a better route than they one they have now. Most people are getting really good pings, but there has to be a better route that can be used to support all users without this.

Edited by Syphr, 13 December 2014 - 06:55 PM.

[LobsterPete]

This is my experience too on TWC, nothing gets past OVH Hosting. The exact spot for me is:
192.99.146.48 - Canada - OVH Hosting

[Kadix]

I was fine until about two hours ago, then...
Microsoft Windows [Version 6.3.9600]
(c) 2013 Microsoft Corporation. All rights reserved.

tracert www.mwomercs.com

Tracing route to www.mwomercs.com [192.99.109.192]
over a maximum of 30 hops:

1 <1 ms 1 ms <1 ms 192.168.0.1
2 20 ms 21 ms 14 ms cpe-173-174-64-1.austin.res.rr.com [173.174.64.1
]
3 22 ms 13 ms 13 ms cpe-173-174-64-1.austin.res.rr.com [173.174.64.1
]
4 8 ms 9 ms 9 ms tge2-3.ausutxla02h.texas.rr.com [66.68.5.97]
5 26 ms 26 ms 25 ms be1.ausutxla04h.texas.rr.com [24.175.41.58]
6 20 ms 13 ms 14 ms be50.ausxtxir02r.texas.rr.com [24.175.43.183]
7 26 ms 18 ms 19 ms agg22.hstqtxl301r.texas.rr.com [24.175.41.48]
8 25 ms 22 ms 27 ms 107.14.17.138
9 19 ms 27 ms 21 ms ae-0-0.cr0.dfw10.tbone.rr.com [66.109.6.39]
10 21 ms 22 ms 20 ms 107.14.19.97
11 26 ms 24 ms 30 ms dls-bb1-link.telia.net [213.248.93.189]
12 77 ms 83 ms 69 ms nyk-bb2-link.telia.net [213.155.130.66]
13 82 ms 76 ms 67 ms nyk-b2-link.telia.net [213.155.130.30]
14 * * * Request timed out.
15 131 ms 118 ms 123 ms 192.99.146.84
16 745 ms 767 ms 759 ms 198.27.73.231
17 124 ms 130 ms 121 ms mwomercs.com [192.99.109.192]

Trace complete.


----------------------

EDIT: My ping times in game WERE in the 500 to 600 ms range a couple of minutes ago, and then dropped back to a "normal" 120ms, even though the tracert still shows slow on the same link.

Edited by Kadix, 15 December 2014 - 07:40 PM.

[NocturnalBeast]

Still having very unstable latency in matches since the last patch. I have done everything I can (without buying new hardware) on my end.

[Clit Beastwood]

Kadix, on 15 December 2014 - 07:38 PM, said:

16 745 ms 767 ms 759 ms 198.27.73.231
Trace complete.

I'm wondering if that hop isn't a router getting overloaded. With the launch of CW, lots of pilots are coming out of retirement.

Ed Steele, on 16 December 2014 - 10:31 AM, said:

Still having very unstable latency in matches since the last patch. I have done everything I can (without buying new hardware) on my end.

Have you tried an SSL vpn?
I use privateinternetaccess for other stuff, it cleaned up mwo a bit, what really helped was a gaming specific VPN - I want WTFast but it doesn't work with MWO, pingzapper does though.

[NocturnalBeast]

Fierostetz, on 16 December 2014 - 10:56 AM, said:

[/size]
Have you tried an SSL vpn?
I use privateinternetaccess for other stuff, it cleaned up mwo a bit, what really helped was a gaming specific VPN - I want WTFast but it doesn't work with MWO, pingzapper does though.

I will look into it when I have time.

[Alekzander Smirnoff]

I'm going to look into pingzapper as well, assuming my wife doesn't beat me for spending $4 for a month to try it out for more than 30 minutes at a time.

[Clit Beastwood]

explain it using math!

"Honey, honey - ok - when I get dc'd or bad lag, I get mad. When I get mad, I drink a few beers. Those beers total 8 bucks. That's 8 bucks each time it freaks out! This is 4 dollars a MONTH. It's already a savings if I only used it once, but I can use it as much as I want... for 4 bucks a month!"

Alekzander Smirnoff, on 16 December 2014 - 02:20 PM, said:

I'm going to look into pingzapper as well, assuming my wife doesn't beat me for spending $4 for a month to try it out for more than 30 minutes at a time.

Can you post back here saying whether it helped you or not? For me, with the trial it was obvious 15 minutes in that it was doing a danged good job so I straight up prepaid a year, on the spot. I don't use it all the time, but when I start hitting lag I fire it up. It's not a game changer, it doesn't confer an advantage, it just un-poops the lag and frequent disconnects that time warner brought me when they "upgraded" my connection. Sure, its 320mbits now, but with fatty packet loss.

[Alekzander Smirnoff]

TWC: "Here's a shiny new gigantic pipe! It has tons of leaks, hell, more packets leak than water going over niagara falls! That's totally ok though, UDP fires those packets like Arnold firing bullets out of his minigun! TCP woes? No worries, we'll strip the headers and convert it all to UDP and fire them even faster!"

Me: "Vegeta, what does the scouter say about his packet loss? IT'S OVER 9000! FML"

[Kadix]

Are we being compensated with additional premium time? Per your initial post, we'll be waiting until January for you to write up what you think is a possible fix. I've got 17 days of premium time active right now.

-------------

• Prior to October 19, my latency was between 80ms to 90ms in game.
  
• On October 20, my latency jumped to 120 to 140ms, which I was able to isolate to a specific hop within your ISP (and provided you with that info). It stabalized to about 120 to 125ms by October 23. Your support staff initially blew me off with a bunch of "torture the customer by asking unrelated questions", and then later admited (in email) that you deliberatly asked my unrelated questions to distract me. Ticket 146005 if you'd like to review.
  
• Last night, my latency spiked to over 700ms for several hours before dropping back to ~120ms.
  
• Tonight, my latency is in the 250ms to 600ms range. Yes, it's that jittery.

I've made ZERO changes to my infrastructure during that time.

I work on an enterprise cloud product with datacenters all over the world. From where I'm sitting in Austin, TX I can get ICMP packets across the public internet (starting at my home Time Warner Cable connection) to webservers in my datacenters in

• Osaka (Japan) in average=155ms min=144ms, max=168ms
  
• Amsterdam (Netherlands) in average=131ms, min=127ms, max=137mm
  
• Ashburn (VA, USA) in average=59ms, min=57ms, max=62ms

Whatever you changed needs to get fixed, and those of us with premium time need to be compensated.

Edited by Kadix, 16 December 2014 - 05:23 PM.

[Clit Beastwood]

Kadix, on 16 December 2014 - 04:39 PM, said:

Are we being compensated with additional premium time since (per your initial post) I'll be waiting until January for a fix?

FWIW, I am *not* using a TWC provided router.

They don't need to compensate anyone for anything, as it's not PGI's fault - PGI helped to identify where the issue lies within TWC's routing - it's up to TWC to fix their own network. PGI can contact TWC and attempt to "recommend" changes, but they cannot affect change on TWC's network. PGI only has control over their internet-facing router on in, it's up to ISP's for everything outside that.

[Macksheen]

I appreciate the focus on TWC.

Another one you should consider looking into in your copious spare time is AT&T Uverse.

I've got both, and honestly TWC has long performed better than AT&T. It's not a constant latency thing; ping tends to be ok. Part of it seems to be DNS related (I swapped to TWC to post this even though I'm farther from the access point because UVerse wasn't able to even load the full web page) - but part is definitely more in-gamey. Lower frame rates is the key symptom.

So, what do you need for me to do to help with this, if anything? I tend to just play on the TWC connection - but I wouldn't be against doing a bit of troubleshooting on the ATT side if it helped others.

[Kadix]

Fierostetz, on 16 December 2014 - 04:47 PM, said:

They don't need to compensate anyone for anything, as it's not PGI's fault - PGI helped to identify where the issue lies within TWC's routing - it's up to TWC to fix their own network. PGI can contact TWC and attempt to "recommend" changes, but they cannot affect change on TWC's network. PGI only has control over their internet-facing router on in, it's up to ISP's for everything outside that.

I would strongly disagree.

Look at the trace I posted that showed a huge problem at the 16th hop. Now, go look up the WHOIS records for the 15th, 16th, and 17th hop.

It's all OVH hosting, which is the company that PGI is paying to host their servers. PGI needs to be working with OVH to resolve the issue.

While it's popular to blame TWC, and I agree TWC does a bunch of sneaky ****, in this case it's not TWC.

[N0MAD]

Kadix, on 16 December 2014 - 06:04 PM, said:

I would strongly disagree.

Look at the trace I posted that showed a huge problem at the 16th hop. Now, go look up the WHOIS records for the 15th, 16th, and 17th hop.

It's all OVH hosting, which is the company that PGI is paying to host their servers. PGI needs to be working with OVH to resolve the issue.

While it's popular to blame TWC, and I agree TWC does a bunch of sneaky ****, in this case it's not TWC.

Its definately OVH,, have same problem, good ping till i hit an OVH server then the 2-3 bounces in OVH space doubles it, they been Ffing with everything and making excuses everywhere but it is OVH, ohh well pay peanuts you get monkeys

[zinetwin]

In the first post they mentioned that PGI is going to try and peer with TWC to help resolve this issue.
But then you remember it's TWC and would likely still screw it up... Man I wish I had another option.
Edit: It does seems stable now though.

Edited by zinetwin, 16 December 2014 - 10:01 PM.

[Clit Beastwood]

Kadix, on 16 December 2014 - 06:04 PM, said:

I would strongly disagree.

Look at the trace I posted that showed a huge problem at the 16th hop. Now, go look up the WHOIS records for the 15th, 16th, and 17th hop.

It's all OVH hosting, which is the company that PGI is paying to host their servers. PGI needs to be working with OVH to resolve the issue.

While it's popular to blame TWC, and I agree TWC does a bunch of sneaky ****, in this case it's not TWC.

Oh I'm not saying that that hop isn't slow, but the *inconsistent* lag is only there on my TWC connection, and from other TWC users I've spoken to they've observed the same. I've tested on several connections, and though my ping dropped pretty hard and my hitreg went to poo with the data center move, the inconsistent ping-bounce is TWC and TWC alone (for me and at least 2-3 others). I'm more than familiar with traceroute, but it only tells a small part of the story. The testing I've done thus far since the jitter started is as follows.

*Tested from a Cox business fiber connection (my office, a few miles from home) - no ping jitter, ping 10ms lower than TWC cable modem.
*Tested from a Cox consumer cable modem (66 miles further south) - no ping jitter, ping 5-10ms less than TWC cable modem.
*Tested from a t-mobile 4g tethering connection (surprisingly fast! - 20ms jitter, ok ping for wireless (120-150ms.
*Tested on my home connection with several routers, but primarily a pfsense firewall running on an i5 with Intel nics.
*Tested all the above with and without SSL vpn, the TWC connection is the *only* one where there was a marked improvement in performance in-game with the VPN enabled.

Note: The last 2 hops in Canada have always been slow for me.. but consistent. For all I know TWC is modifying the packet headers or something, I haven't had time to pursue it (holidays and all) further. When I have more time I'd like to try psping or something like it to test pings to specific ports, but thats going to have to wait until after Christmas.

Edited by Fierostetz, 17 December 2014 - 09:40 AM.

[Alekzander Smirnoff]

I've been using ping plotter to keep an eye on the route from my location to theirs and I can say there are a few instances where I get no response (ICMP requests may be ignored) or its just way the **** out there. Need to get my lazy but downloading pingzapper and try it out one night when I usually experience "disconnected from game server" 8000 times.

[Alekzander Smirnoff]

OK, so I installed pingzapper, my latency is still the same around 40 ms with it on or off. Now my other question is, I'm in Kentucky, and the servers are (I'm assuming here) in Vancouver or near by there. I'm going to also make the assumption that you assign the server setting to where the server actually is, because I've not read anything about how to set that one. Right now it funneled my traffic out of New York, NY and whatever route it took from there I was still getting "disconnected from game server." I'm going to flip it over to West which dumps it out in Los Angelas, CA and see what happens.