Low ping from cmd prompt high ping in game

[BoPilot]

I have been having this problem on and off for a while now can anyone make sense of this?

If the pic above doesn't show I also attached it to the post, sorry i'm not good with message boards.

As you can see in the pic a cmd prompt ping to google dns avgs 40-60ms (it was fluctuating more than usual my GF was streaming video at the time of the test). When there is no activity it levels off around 35-40ms with small 1-2ms fluctuations. If you run a ping test through any variety of "ping/latency test" sites it will produce a number from 150-400ms. The closest servers on the steam game server browser show about 160ms.

Now this only happens from time to time and when this is not happening the ping times to google dns remain the same and will for the most part match the ping results in the ping tests from the various websites in the image. I have tried a different router and the issue remains. It's driving me bananas and any insight would be appreciated.

 

[Mussels]

the jittery pings to google are a sign of problems as well, they should be fairly stable.

can you ping your routers IP as well and see how it is behaving? right now it looks like a network thats being saturated, such as from downloads/uploads/torrents.

 

[BoPilot]

the jittery pings to google are a sign of problems as well, they should be fairly stable.

can you ping your routers IP as well and see how it is behaving? right now it looks like a network thats being saturated, such as from downloads/uploads/torrents.

My GF was streaming video while i ran that ping test causing the fluctuation it usually only fluctuates 1-2ms at most. The problem still exists regardless of saturation and when its not acting up I can get sub 100ms ping time across the board even during heavy saturation. I monitor the router's bandwidth monitor so i know when its under heavy use. Ping to the router is steady 1ms.

 

[Mussels]

My GF was streaming video while i ran that ping test causing the fluctuation it usually only fluctuates 1-2ms at most. The problem still exists regardless of saturation and when its not acting up I can get sub 100ms ping time across the board even during heavy saturation. I monitor the router's bandwidth monitor so i know when its under heavy use. Ping to the router is steady 1ms.

the video streaming is simply eating your bandwidth. you basically need to cap the internet bandwidth somehow.

netflix here causes the same issue (as does youtube) so i've got a bandwidth shaping system in my router to keep every IP at <50% of my bandwidth (approx 500KB/s) so that she can do what she wants on any one device without it lagging everything out.

 

[BoPilot]

the video streaming is simply eating your bandwidth. you basically need to cap the internet bandwidth somehow.

netflix here causes the same issue (as does youtube) so i've got a bandwidth shaping system in my router to keep every IP at <50% of my bandwidth (approx 500KB/s) so that she can do what she wants on any one device without it lagging everything out.

Maybe I was not being clear I still have this problem when the internet is not in use by anyone, there is nothing consuming bandwidth but my own tests (confirmed via router WAN bandwidth monitor) and this problem still exists. The image is only to show that there is a 120-300ms differences between the cmd prompt ping and the ping results on tests. This still happens when no one is steaming or consuming bandwidth only difference is the fluctuations in the google dns ping are much smaller.

 

[Mussels]

Maybe I was not being clear I still have this problem when the internet is not in use by anyone, there is nothing consuming bandwidth but my own tests and this problem still exists.

then your connection is unstable from the modem outwards. if you're on DSL you can try another modem/another phone cable, if seriously no bandwidth is in use then it may be on your ISP's end.

 

[BoPilot]

then your connection is unstable from the modem outwards. if you're on DSL you can try another modem/another phone cable, if seriously no bandwidth is in use then it may be on your ISP's end.

Do you have any idea why there would be such a large difference between cmd prompt ping command and the website tests? I know ping command uses ICMP and its a bit faster because of less overhead but that doesn't explain the huge difference in latency. Also don't you think that if the modem/ISP had an issue that the pings from cmd prompt would be equally slow?

 

[BoPilot]

Now the ping times are normal again 35ms-38ms cmd prompt 41ms from pingtest site. Just tested it 20min ago with 0 bandwidth usage on the router's monitor and got 35-38ms cmd prompt 160-350ms from 3 different ping test sites. As far as I can tell its completely random, the cmd prompt latency numbers are always same.

I can live with this, I have for a long time now. I just want to make some sense of it.

 

[Mussels]

Do you have any idea why there would be such a large difference between cmd prompt ping command and the website tests? I know ping command uses ICMP and its a bit faster because of less overhead but that doesn't explain the huge difference in latency. Also don't you think that if the modem/ISP had an issue that the pings from cmd prompt would be equally slow?

because those website tests use more bandwidth. your problem is that the more bandwidth is used, the worse it gets.

 

[Mussels]

you asked why it gets worse on the larger tests, that's why. they're further away and use more data (they're loading the website itself too, remember)

again i repeat: the more you use the worse it gets, and its your modem outwards. The fact your pings vary so wildly is a sign of some kind of instability. I cant guess any more without information from you.

 

[BoPilot]

you asked why it gets worse on the larger tests, that's why. they're further away and use more data (they're loading the website itself too, remember)

again i repeat: the more you use the worse it gets, and its your modem outwards. The fact your pings vary so wildly is a sign of some kind of instability. I cant guess any more without information from you.

My firewall show the website ping test using less than 2KBps so that isn't the case. If you missed my last post the ping times have normalized again with the cmd prompt pings matching the website test results. A test was run not 20min earlier with the router bandwidth monitor open showing no bandwidth usage besides the 2KBps for the test and the results averaged 160-350ms. Just for reference since I am not having the problem at this moment I went ahead and saturated my downstream to see how high my pings get and they avg 70-100ms under full saturation of the downstream.

Considering I can get 70ms ping times under full saturation when I'm not having the problem and my pings are 160-350ms under 0 load when the problem is occurring I have to assume this is not a bandwidth issue. If I always got 160-350ms in the website tests I would agree with you but this is only the case when I am having this issue which seemingly comes and goes at random regardless of bandwidth usage.

 

[95Viper]

if you're on DSL you can try another modem/another phone cable, if seriously no bandwidth is in use then it may be on your ISP's end.

Alyrica Internet is completely wireless to your home, meaning that if you have a computer, electrical power, and line of sight to one of our towers, you can use our service.

@OP:

Have you tried a tracert to the website you are trying to get?

 

[BoPilot]

Alyrica Internet is completely wireless to your home, meaning that if you have a computer, electrical power, and line of sight to one of our towers, you can use our service.

@OP:

Have you tried a tracert to the website you are trying to get?

I have not but at the current moment the problem has vanished as it does randomly. I will run a tracert now to all 3 test sites I used and when the problem resurfaces I will run another tracert to the 3 and report back.

 

[BoPilot]

Alyrica Internet is completely wireless to your home, meaning that if you have a computer, electrical power, and line of sight to one of our towers, you can use our service.

@OP:

Have you tried a tracert to the website you are trying to get?

The top prompt is tracert when the problem is not occurring.

 

[Aquinus]

The top prompt is tracert when the problem is not occurring.

See the jump in ping on hop #3? (e01.kattare-gw.cvo.alyrica.net [206.192.248.197]). That's your problem. That's an indication that your connection upstream to the DSL provider is bad. You're probably a long way from the office or you there is a telephone run that's on the fritz (water got in the cable, weak connection, old cable that has weakened over time, etc.) This feels like a signal strength/SNR issue with your DSL modem otherwise it wouldn't show up on the very first hop outside of your own router (first non-private IP address.) It's worth noting that two private internal IPs is a little strange. (192.168.1.1 to 10.58.98.1.)

 

[BoPilot]

See the jump in ping on hop #3? (e01.kattare-gw.cvo.alyrica.net [206.192.248.197]). That's your problem. That's an indication that your connection upstream to the DSL provider is bad. You're probably a long way from the office or you there is a telephone run that's on the fritz (water got in the cable, weak connection, old cable that has weakened over time, etc.) This feels like a signal strength/SNR issue with your DSL modem otherwise it wouldn't show up on the very first hop outside of your own router (first non-private IP address.) It's worth noting that two private internal IPs is a little strange. (192.168.1.1 to 10.58.98.1.)

No DSL here I use a WISP called Alyrica so there is no modem. Does that one hop there really justify a jump of 300+ms in my ping? Doesn't seem that much higher than the top tracert and other hops seems to have an even larger difference between the 2 tests.

 

[Aquinus]

No DSL here I use a WISP called Alyrica so there is no modem. Does that one hop there really justify a jump of 300+ms in my ping? Doesn't seem that much higher than the top tracert and other hops seems to have an even larger difference between the 2 tests.

ICMP packets are much different in nature than TCP packets, much faster turn around, no handshake, no acknowledgement other than the ICMP ping response. I thought I read that you had DSL, my mistake. Wireless can be susceptible to any number of issues as well. Any obstruction can cause issues which even includes rain. A lot of that could be out of your control.

 

[BoPilot]

ICMP packets are much different in nature than TCP packets, much faster turn around, no handshake, no acknowledgement other than the ICMP ping response. I thought I read that you had DSL, my mistake. Wireless can be susceptible to any number of issues as well. Any obstruction can cause issues which even includes rain. A lot of that could be out of your control.

I understand that TCP packets have much more overhead but that doesn't explain how when I am having this issue my TCP ping tests show 300+ms and when I am not having the issue the TCP ping tests show 40ms the same as a cmd prompt ping to google DNS. If it was my wireless connection issue to the ISP don't you think it would affect the cmd prompt times as well or perhaps the amount of available bandwidth? When the issue is present the cmd prompt ping times do not change and full bandwidth is still available, I just get bad ping times with TCP connections.

Thanks for all the help btw, sorry if I ask too many questions I really just want to get to the core of the issue.

 

[Aquinus]

I understand that TCP packets have much more overhead but that doesn't explain how when I am having this issue my TCP ping tests show 300+ms and when I am not having the issue the TCP ping tests show 40ms the same as a cmd prompt ping to google DNS. If it was my wireless connection issue to the ISP don't you think it would affect the cmd prompt times as well or perhaps the amount of available bandwidth? When the issue is present the cmd prompt ping times do not change and full bandwidth is still available, I just get bad ping times with TCP connections.

Thanks for all the help btw, sorry if I ask too many questions I really just want to get to the core of the issue.

ICMP is inherently faster because only one message goes out and one message comes back. With TCP, one message gets sent, the receiving server then has to send an acknowledgement that the message was received, then the server again sends a packet back to the client where the client is then responsible to send back an acknowledgement to say the message was received. So TCP actually sends twice as many messages to do the same job as ICMP. Mainly because that's how TCP works. UDP on the other hand is much more like ICMP where you're just sending packets out without any expectation that they worked, are received in order, or even received which makes it inherently faster as well.

Simply put, I think it's just because it's wireless. So many conditions can dictate how good a WISP is going to be. One this is for sure, it's not likely to give you a consistent ping like a physical connection typically should.

 

[BoPilot]

ICMP is inherently faster because only one message goes out and one message comes back. With TCP, one message gets sent, the receiving server then has to send an acknowledgement that the message was received, then the server again sends a packet back to the client where the client is then responsible to send back an acknowledgement to say the message was received. So TCP actually sends twice as many messages to do the same job as ICMP. Mainly because that's how TCP works. UDP on the other hand is much more like ICMP where you're just sending packets out without any expectation that they worked, are received in order, or even received which makes it inherently faster as well.

Simply put, I think it's just because it's wireless. So many conditions can dictate how good a WISP is going to be. One this is for sure, it's not likely to give you a consistent ping like a physical connection typically should.

I would not disagree that it could be a issue with my ISPs wireless system, but I can't agree that it is simply because it is a different protocal, the additional overhead of TCP would not add 300ms of latency. I have proven this by running the TCP tests when I am not having the problem, in these cases the ICMP ping and the TCP ping are the exact same with maybe 1ms difference not 300ms. However if my wireless connection is the issue why would this only affect TCP connections and not ICMP connections? In every wireless connection issue I have seen the ICMP pings are equally affected.

 

[95Viper]

Here is a real good read... and, if you understand this you can do some good troubleshooting: A Practical Guide to (Correctly) Troubleshooting with Traceroute

Me, I am going to take a guess and say you are having the same problems with Alter.net, that we had on the east coast with them.
Alter.net's equipment was causing timeouts, high latency, lost packets, etc.... we (when I say we, here, I am talking about the Fios community) have continuously complained about the problem and now it seems to be better, if not fixed.
Ain't to much you can do, except, complain to your isp and hope they can resolve the issue.

You can try this diagnostic test at MLab's site:
NDT (Network Diagnostic Test)

It is an informative test, especially, the debug and detail sections.
INFO: NDT may, or may not run, depending on your JAVA version, or, if you have Websockets or not.

Also, run the tests with, and without, your A/V software and firewall.
You, may, want to turn off any other, so-called, internet improvement software you may be running, if any.

 

[BoPilot]

Here is a real good read... and, if you understand this you can do some good troubleshooting: A Practical Guide to (Correctly) Troubleshooting with Traceroute

Me, I am going to take a guess and say you are having the same problems with Alter.net, that we had on the east coast with them.
Alter.net's equipment was causing timeouts, high latency, lost packets, etc.... we (when I say we, here, I am talking about the Fios community) have continuously complained about the problem and now it seems to be better, if not fixed.
Ain't to much you can do, except, complain to your isp and hope they can resolve the issue.

You can try this diagnostic test at MLab's site:
NDT (Network Diagnostic Test)

It is an informative test, especially, the debug and detail sections.
INFO: NDT may, or may not run, depending on your JAVA version, or, if you have Websockets or not.

Also, run the tests with, and without, your A/V software and firewall.
You, may, want to turn off any other, so-called, internet improvement software you may be running, if any.

Thanks for the insight on Alter.net I will contract Alyrica with all the information gathered and see if they can provide any insight. I will also get to reading that guide and trying out the diagnostic test as soon as I get off work. Thanks again for the help 95Viper!