Jump to content

Lags

Sebast1an

By Sebast1an
in Help

 Share

Recommended Posts

Sebast1an Collaborator

Sebast1an

Posted
Posted (edited)

Heyo,

recently on ETL1 my ping has been 70-90 whereas normally it used to be 50-60. On top of this some maps just lag more than before - I've noticed it quite well on Erdenberg as that map lags a lot these days compared to before (lag as in spikes in lagometer, fps is usually fine). On my end nothing has changed in the last year so I guess this could be related to server setting changes somehow? I tried ETL2 and ETLC, with bots, and on both servers my ping was 48.

Only server change I can recall that could have an effect is /rate was forced to 35000 instead of 25000 some time ago. Not sure if it could have this kind of effect though. I've tried all sort of tweaks to my cfg without much success but I hope some of you may have some new suggestions to try. :) Cheers.

I'm rocking a whopping 24/1 Mbps connection which hasn't changed in a while. Let me know if you need to know some other specs. Also here's my config:

  Reveal hidden contents

 

Edited by Sebast1an
Added the cl_maxpackets value that was missing from the cfg for some reason
  • Founders
MiniMuppet Grand Master

MiniMuppet

Posted
  • Founders
Posted

Other then the rate setting we haven't changed anything. Other then adding new builds. I can lower the rate setting next time you are on to see if it makes a difference but not sure it would.

Sebast1an Collaborator

Sebast1an

Posted
  • Author
Posted

Thank you Mini, I think it's a good idea to try regardless just so we know it's not about that.

Just tested it on ETL1 with bots, thinking it might have something to do with the amount of people. But unfortunately that makes no difference.

Zyxpix Rookie

Zyxpix

Posted
Posted

This is a routing problem I think and you should contact your ISP

ETLegacy main is in France

ETLegacy 2 is in Germany

  • Founders
MiniMuppet Grand Master

MiniMuppet

Posted
  • Founders
Posted
  On 9/5/2021 at 5:06 AM, Sebast1an said:

Thank you Mini, I think it's a good idea to try regardless just so we know it's not about that.

Just tested it on ETL1 with bots, thinking it might have something to do with the amount of people. But unfortunately that makes no difference.

Expand  

I am on nights until wed will try it then.

Sebast1an Collaborator

Sebast1an

Posted
  • Author
Posted
  On 9/5/2021 at 7:14 AM, Zyxpix said:

This is a routing problem I think and you should contact your ISP

ETLegacy main is in France

ETLegacy 2 is in Germany

Expand  

A routing problem sounds very logical to me, thank you. I did tracert on both ETL1 and ETL2 and these were the results:

ETL1:

tracert  37.187.79.49

Tracing route to ns3367630.ip-37-187-79.eu [37.187.79.49]
over a maximum of 30 hops:

  1     6 ms     5 ms     5 ms  *myipremoved*
  2     *        8 ms     7 ms  141.208.186.161
  3     8 ms     9 ms     8 ms  141.208.186.162
  4     8 ms     8 ms     9 ms  141.208.186.163
  5    11 ms    10 ms    10 ms  141.208.195.149
  6    12 ms    11 ms    11 ms  62.115.143.50
  7     *        *        *     Request timed out.
  8    37 ms    37 ms    38 ms  ffm-bb1-link.ip.twelve99.net [62.115.143.29]
  9    38 ms    38 ms    37 ms  ffm-b11-link.ip.twelve99.net [62.115.124.117]
 10    40 ms    40 ms    41 ms  be100-163.fra-5-a9.de.eu [178.33.100.250]
 11     *        *        *     Request timed out.
 12     *        *        *     Request timed out.
 13     *        *        *     Request timed out.
 14    52 ms    52 ms    51 ms  be105.gra-g2-nc5.fr.eu [91.121.131.18]
 15     *        *        *     Request timed out.
 16     *        *        *     Request timed out.
 17     *        *        *     Request timed out.
 18    53 ms    53 ms    52 ms  ns3367630.ip-37-187-79.eu [37.187.79.49]

Trace complete.

 

ETL2:

tracert 46.4.80.110

Tracing route to static.110.80.4.46.clients.your-server.de [46.4.80.110]
over a maximum of 30 hops:

  1     6 ms     5 ms     5 ms  *myipremoved*
  2     8 ms     7 ms     7 ms  141.208.186.161
  3     8 ms     7 ms     8 ms  141.208.186.162
  4     8 ms     8 ms     8 ms  141.208.186.163
  5     *       11 ms    11 ms  141.208.195.149
  6    11 ms    11 ms    11 ms  62.115.143.50
  7    37 ms    38 ms    38 ms  s-bb2-link.ip.twelve99.net [62.115.123.202]
  8    36 ms    35 ms    36 ms  ffm-bb2-link.ip.twelve99.net [62.115.138.105]
  9    38 ms    38 ms    37 ms  ffm-b5-link.ip.twelve99.net [62.115.114.91]
 10    47 ms    42 ms    42 ms  hetzner-ic326013-ffm-b5.ip.twelve99-cust.net [213.248.70.3]
 11   193 ms     *       47 ms  core23.fsn1.hetzner.com [213.239.224.249]
 12    43 ms    43 ms    45 ms  ex9k1.dc5.fsn1.hetzner.com [213.239.229.122]
 13    43 ms    43 ms    44 ms  static.110.80.4.46.clients.your-server.de [46.4.80.110]

Trace complete.

 

Could those "request timed out" messages on ETL1 be the cause of the problem?

Zyxpix Rookie

Zyxpix

Posted
Posted
  On 9/5/2021 at 3:20 PM, Sebast1an said:

Could those "request timed out" messages on ETL1 be the cause of the problem?

Expand  

No, I have the same thing : 

Détermination de l’itinéraire vers ns3367630.ip-37-187-79.eu [37.187.79.49]
avec un maximum de 30 sauts :

  1    <1 ms    <1 ms    <1 ms  mymodem [192.168.1.1]
  2     5 ms     5 ms     5 ms  10.24.1.51
  3     *        *        *     Délai d’attente de la demande dépassé.
  4     5 ms     5 ms     5 ms  ae-60-100.ibrstr5.isp.proximus.be [91.183.241.168]
  5     *        *        5 ms  lag-30-1000.ibrmar5.isp.proximus.be [91.183.246.125]
  6     *        *        *     Délai d’attente de la demande dépassé.
  7     *        *        *     Délai d’attente de la demande dépassé.
  8     *        *        *     Délai d’attente de la demande dépassé.
  9     *        *        *     Délai d’attente de la demande dépassé.
 10     *        *        *     Délai d’attente de la demande dépassé.
 11     *        *        *     Délai d’attente de la demande dépassé.
 12     9 ms     8 ms     8 ms  ns3367630.ip-37-187-79.eu [37.187.79.49]

Itinéraire déterminé.

It just means that ICMP request are blocked, nothing more.

StarScream Enthusiast

StarScream

Posted
Posted
  On 9/5/2021 at 3:20 PM, Sebast1an said:

37 ms    37 ms    38 ms  ffm-bb1-link.ip.twelve99.net [62.115.143.29]

Expand  

So, yeah, it looks like your problems start when you hit Telia's network. This doesn't surprise me, I've had issues in my work with them and they are singularly unhelpful.

 

  On 9/5/2021 at 3:20 PM, Sebast1an said:

10    47 ms    42 ms    42 ms  hetzner-ic326013-ffm-b5.ip.twelve99-cust.net [213.248.70.3]
 11   193 ms     *       47 ms  core23.fsn1.hetzner.com [213.239.224.249]

Expand  

This might be a blip or a problem peering between Telia (backbone) and Hetzner (the host). Given that you get +20ms to both servers when you hit Telia too, I would say that Telia is having peering problems.

As an end user there's very little you can do, just let your ISP know that there seem to be issues with Telia.

 

 

  • Founders
MiniMuppet Grand Master

MiniMuppet

Posted
  • Founders
Posted

to be honest they both look good and it just skips the ones it can't connect to. If there was a issue your pings would be a lot higher.

StiffWrists Apprentice

StiffWrists

Posted
Posted (edited)
  On 9/4/2021 at 8:10 PM, Sebast1an said:

On top of this some maps just lag more than before - I've noticed it quite well on Erdenberg as that map lags a lot these days compared to before (lag as in spikes in lagometer, fps is usually fine)

Expand  

Some maps lag way worse than others from my experience and Erdenberg is absolutely the worst culprit. It's impossible for me to get a stable 200ms ping there. I have also noticed that the latency spikes are aggravated the more players there are. My theory has been that it's due to two things:

1. Bandwidth bottleneck somewhere in the network path. Bandwidth saturation produces knock-on effects like elevated or unstable latencies.
2. Map geometry is not well sectioned (or has leaks) so you end up with far more server traffic than necessary to update invisible players.

(2) is maybe testable by recording a demo and playing back using r_showNormals 1.

From the EU, network bottlenecks really shouldn't be a problem. To check this, you could maybe run Speedtest (https://www.speedtest.net/) from your computer and picking the ISP the main ETL1 server is on.

One misconception behind network issues is that the problem lies on one end or the other i.e. the ISP or the game server in this case. In reality, it very much depends on the routing between the two endpoints and the problematic node on the path could belong to an intermediary.

As far as client tweaks is concerned, there is the setting of packet duplication to off (cl_packetDup 0), reducing snaps, reducing rate and reducing cl_maxPackets. Except for cl_packetDup, the ETL1 server is unfortunately too restrictive for any of the above to work. I have always wondered if it would be better to set a range of acceptable snap and rate values rather than fixing it to 40 and 35000 respectively.

On a sidenote: regarding traceroutes (or tracert on Windows), you have to be careful with interpretation.

1. Most service providers will ask you to provide traces in both directions because the forward and reverse paths may not be the same.

2. It is alright for intermediate nodes to not respond to traces. Pings and Traceroutes use a different protocol (ICMP) and are considered 'control traffic' instead of 'data traffic' and switches/routers prioritize forwarding data. The right way to interpret Traceroute results is to look at the trace in reverse, beginning from the destination. Note down the latency and losses, then work backwards to the earlier nodes to observe the progression.

Edited by StiffWrists
Fix mistakes
  • Like 1
lost Collaborator

lost

Posted
Posted

For me on non standard map like enderberg the problem was with com_hunkmegs and com_soundmegs, values were to low, when i adjusted them everything started to working perfect

Mister J Rising Star

Mister J

Posted
Posted

I increased all those megs too, problems persisted. I did notice that ET is using much more resources than it used to. Not that this should be a problem in 2021. My wooden computer however is actually starting to have issues running the game. :biggrin:

Sebast1an Collaborator

Sebast1an

Posted
  • Author
Posted
  On 9/14/2021 at 11:13 AM, lost said:

For me on non standard map like enderberg the problem was with com_hunkmegs and com_soundmegs, values were to low, when i adjusted them everything started to working perfect

Expand  

Which values do you use for hunkmegs, soundmegs and zonemegs?

StiffWrists Apprentice

StiffWrists

Posted
Posted

I'm basically right about something wrong with Erdenberg. Every area is 'visible' no matter where you are. In the screenshots attached, you can see that both radars are drawn despite how far and how obscured they are from the Allied first spawn. Data for practically every player in the map will be transmitted no matter where they are, generating a lot of unnecessary traffic.

Compare it with Supply Depot, where most of the hidden areas are dropped from processing.

Is there an updated version of this map that fixes this issue?

2021-09-17-022005-erdenberg_t1.jpg

2021-09-17-022516-supply.jpg

  • Founders
MrMuppet Grand Master

MrMuppet

Posted
  • Founders
Posted

Unfortunately I believe the version we run is the last version released, such a shame because it is a very popular map on the server, @kemon is this in your workstream?

  • 2 weeks later...
lost Collaborator

lost

Posted
Posted

Hello

From yesterday something strange happened on the main server. I had all the time ping around 50 and right now periodically it is jumping to about 80 and lowering again to 50. This is not problem with my connection because I checked it. I checked if this happening on the other server and i dont have any problem on other servers from TM. It does not depend on the number of players. It is happening all the time right now. Can someone restart server or something?

  • Haha 1

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

×
 Share
Go to topic listing
×
×
  • Create New...