Servers Lag

[Muerte]

What connection you have? dsl?

Looks like your home2telecom installation is bad or your connection is on interleave path, faulty cables, splitter or router.

[2d]

I have comcast cable

what  the hell is a home2telercom?

[Muerte]

Then rats, cats or dogs eating your cable
or just your isp oversold connections so u have local lags...

[ex]

or just your isp oversold connections so u have local lags...

This is the most realistic scenario.  Demand for internet is far outpacing the technicians' ability to keep up with current needs (or in actuality, there aren't enough qualified technicians to keep up with demand for installing high speed internet).  This bullshit will go away when gigabit technology becomes the norm, which is coming quickly.  Google Fiber isn't the only player on the field.  Eventually, Google will force all the prices down and everyone will be dealing with faster internet with less bullshit, and this will also force the cable companies to bump their speeds up to gigabit levels.  Just recently 3 gbps was achieved on a Comcast coaxial internet connection.  Current DOCSIS 3.0 modems can handle over 1.2 gbps comfortably with a 32 channel setup - not sure what they used to get 3 gbps - maybe a DOCSIS 3.1 prototype.  ;>

I just realized I drifted off topic.  I am just kind of amazed at the way they overcharge for internet services.  Time Warner charges around $70 for 15 mbps internet in the Dallas area (about to be bumped to 50 mbps soon for same price).  If Google Fiber was here, 1 gbps (1000 mbps) would cost me the exact same.  Just mind boggling that we're forced to pay so damn much for sub-par internet.  Greed is a bad thing for the interwebs.

BTW, consider getting off of Comcast.  I believe there are data usage "caps"/taxes that happen once you crest past 300 Gb depending on your area.  Basically, use more than that on your home-based internet, you get to pay an upcharge for every 50 Gb thereon.

Here's some good reading on the topic of caps:

http://stopthecap.com/tag/time-warner-cable/

And here's something specific about the types of caps different providers have in place, at least as of 2013:

https://gigaom.com/2013/11/15/data-cap-2013/

What's good about Time Warner is that the internet is still truly unlimited and uncapped, and the company is planning gigabit speeds by year-end 2016.  And if the current trend with the Maxx markets are any clue, which turn all of their speeds up anywhere from 3 to 6 times over (15 to 50, 20 to 100, 30 to 200, 50 to 300), then it's not impossible to think about the future of the internet.

DOCSIS 3.1 teaser:

http://www.multichannel.com/news/news-articles/docsis-31-speeds-ahead/374179

10 gbps internet, anyone?  :>

[quadz]

Im not sure how to judge my connection overall. I am losing on average 30% of my packets sent and received.... but then again it says ts is receiving 100%.... i dont know how that makes sense...

Again, the ONLY part that matters to Quake2 is the final hop at tastyspleen.net, which is showing 0% loss.

i guess this program implies that i use the same path for travel to and from ts? so the loss is inclusive for both directions of travel?

No, however I can run a trace from the server -> you to find the route in that direction.

Code: [Select]

Host                                                 Loss%   Snt   Last   Avg  Best  Wrst StDev
 1. 67.228.69.113-static.reverse.softlayer.com         0.0%  1154    0.3   0.3   0.2  71.9   3.1
 2. ae13.dar02.sr01.dal01.networklayer.com             0.0%  1154    0.2   0.3   0.2  25.5   2.1
 3. ae6.bbr02.eq01.dal03.networklayer.com              0.0%  1154    0.2   0.2   0.2  70.7   2.8
 4. be-112-pe03.1950stemmons.tx.ibone.comcast.net      0.0%  1154    0.7   0.7   0.6  33.7   1.6
 5. he-0-6-0-0-cr01.dallas.tx.ibone.comcast.net        0.0%  1154    2.3   2.2   0.5   5.1   1.2
    he-1-2-0-0-cr01.dallas.tx.ibone.comcast.net
    he-2-3-0-0-cr01.dallas.tx.ibone.comcast.net
    he-2-2-0-0-cr01.dallas.tx.ibone.comcast.net
    he-1-3-0-0-cr01.dallas.tx.ibone.comcast.net
    he-0-5-0-0-cr01.dallas.tx.ibone.comcast.net
    he-1-4-0-0-cr01.dallas.tx.ibone.comcast.net
 6. be-11315-cr02.losangeles.ca.ibone.comcast.net      0.0%  1154   34.3  34.1  33.2  39.9   0.7
 7. be-7922-ar01.hayward.ca.sfba.comcast.net           0.0%  1154   43.4  43.3  43.2  46.5   0.3
 8. he-0-8-0-0-ar01.oakland.ca.sfba.comcast.net        0.3%  1154   42.3  43.5  40.8  47.8   1.6
    he-0-14-0-0-ar01.oakland.ca.sfba.comcast.net
    he-0-9-0-0-ar01.oakland.ca.sfba.comcast.net
    he-0-15-0-0-ar01.oakland.ca.sfba.comcast.net
 9. te-0-0-0-0-sur04.sanrafael.ca.sfba.comcast.net     0.0%  1154   46.4  46.2  46.1  48.3   0.4
10. te-7-0-acr03.sanrafael.ca.sfba.comcast.net         0.0%  1154   43.5  44.5  43.4  50.0   1.3
    te-7-1-acr03.sanrafael.ca.sfba.comcast.net
11. ???

Your local router isn't answering pings, so stats are unavailable for the final hop.

But the rest of the route is very clean. (The slight loss shown at hop 8 would only be relevant if it carried over to subsequent hops, which it doesn't appear to.)

I suspect something is broken with WinMTR when run from your computer, as its results are not making much sense.

In any case, the route from tastyspleen.net to you looks good.

[|iR|Focalor]

or just your isp oversold connections so u have local lags...

This is the most realistic scenario.  Demand for internet is far outpacing the technicians' ability to keep up with current needs (or in actuality, there aren't enough qualified technicians to keep up with demand for installing high speed internet).  This bullshit will go away when gigabit technology becomes the norm, which is coming quickly.

I'm no expert on the subject, not even a novice, but I disagree that their supply is falling behind the demand. I think these ISPs have PLENTY of bandwidth people could be using, they just throttle it back, puke excuses in our ears, and overcharge for slower service than what they could provide if they wanted to simply because... they can. Who's gonna stop them? They're in business to make money, and they're gonna squeeze and wring people out as much as they can. I'd bet my right eye they have people they specifically hire to study the markets and their customers and figure out what the figures are they can (over)charge people to maximize profits without (over)charging too much and losing customers because they can't afford the service at that price.

I wouldn't put it past all these telecom companies in the US to be collectively involved in corporate racketeering and price fixing either. Lots of businesses do the same shit. It's illegal as hell, but they do it anyway.

[2d]

I suspect something is broken with WinMTR when run from your computer, as its results are not making much sense.

rofl...

just my luck, the program dies with me, i got the midas touch

I guess my computer is just outdated or dying. I AM running an emachine, but then again q2 is a 20 yr old game...
 
maybe the video card is shot. oh well thx for the help 

[Muerte]

Again, the ONLY part that matters to Quake2 is the final hop at tastyspleen.net, which is showing 0% loss.

After
6 years on DialUp
2 years on crap cable connection
~3 years on crap ADSL
~3 years on kinda decent ADSL
1 year on ADSL interleaved (interleave depth 256)
1 year on mobile net
2 years on wifi over adsl (now)

I would say everything is matter in Quake2
- local lag (oversold cable connections by ISP)
- lowered upload speed (by ISP without notification)
- lower ADSL MTU by ISP (now max packets is lower then native Quake2 packet...)
- packet loss on route
- now wifi 'signal' 'increased' with 2 beer can  (distance is about 250-300 yards)     shared with 2pc 1 laptop 3 android phones and me 'playing' Quake2 on 4/0.8 mbit ADSL

No, however I can run a trace from the server -> you to find the route in that direction.
Your local router isn't answering pings, so stats are unavailable for the final hop.
But the rest of the route is very clean. (The slight loss shown at hop 8 would only be relevant if it carried over to subsequent hops, which it doesn't appear to.)
I suspect something is broken with WinMTR when run from your computer, as its results are not making much sense.

In any case, the route from tastyspleen.net to you looks good.

Judging by his 1st hops its clear he have local problem
- local installation
- splitter (whatever he have connected on that cable)
- local hub
- router problem

You should try ping/trace route his IP address when he is on line (on forum) (WinMTR don't show router IP) i guess he have dynamic IP

for example my WinMTR in attachment
I know my 'local' problem visible on this winMTR report (even with rail on 20ish ping i must shot at front of player to hit him (not just because local lag i have also packet loss on hops to nearest server because bad local routes)

[quadz]

I would say everything is matter in Quake2

I'm not talking about opinions and feelings. I'm referring to round trip stats measured between his computer and tastyspleen.net as displayed in those screenshots.

By definition, none of the other hops along the route had packets with a high enough TTL to reach tastyspleen. None of the stats for the other hops (strange as they were) reflect what is happening when a packet has high enough TTL to reach tastyspleen.

Even on the weird WinMTR traces, the packets that are actually reaching tastyspleen.net--i.e. those not having a low TTL that expires in transit--are showing 0% packetloss and good average latency. (What's not clear is how much variation/jitter in latency may exist. WinMTR lacks the std. deviation column.)

Judging by his 1st hops its clear he have local problem

This claim not supported by evidence, and is contradicted by the perfect final hop.

The most one could say is the strange WinMTR data should be thrown out as unreliable. The fact of the perfect final hop is evidence that the strange numbers in the rest of the route do not reflect reality.

Further, the reverse trace from tastyspleen back to (just before) his router, also displays a clean path, displaying none of the nonsense from the WinMTR trace. (Further lending weight to that data being spurious.)

All the evidence so far points to two things:

1. If a packet's TTL is too low so that it expires in transit before reaching tastyspleen.net, then he's getting weird numbers.
2. If a packet's TTL is high enough to actually reach tastyspleen.net (The ONLY scenario in which packets could actually be reaching the Quake2 server) then there's 0% packetloss and good average latency.

That is all the data is showing.


Would it help to explain in more detail how traceroute works?





P.S.

for example my WinMTR in attachment
I know my 'local' problem visible on this winMTR report

Your situation is different. Your packetloss is consistently reflected in subsequent hops, all the way to tastyspleen.net, which indicates a genuine (non-TTL-based) issue earlier in your route.

His stats don't show that.

[Muerte]

You convinced me so no need for further explanation

I'm using PingPlotter every day to find "playable server"

Usual connection 2 local quake server few miles from me in attachment 

[2d]

rofl

i think i figured out most of why i am lagging

i was watching task manager while playing and audiodg.exe kept jumping like a rabbit

so i shut the process down and lost audio for a while, and my gameplay significantly improved....

so i went into quake options to see my audio config

and to my surprise, enable cd music was on, direct writing was on, and highest sound quality was on.....

i disabled those things with the quickness.....

improvement indeed 

[Sgt. Dick]

Then rats, cats or dogs eating your cable.

Vae is the only person I ever heard of that happening to