Netgear XR700 High packet loss spikes!!!

[jpoison]

Anyone with this issue ?

[Netduma Jack]

32 minutes ago, jpoison said:

Anyone with this issue ?

Hi, welcome to the forum! Packet Loss isn't usually caused by routers, rather it tends to be upstream equipment like your Ethernet cables, Modem or line to blame.

- Are you testing on a wired connection?
- How are you testing? Is this on Pingplotter?
- Remove the XR700 from your setup. Do you still get Packet Loss?

[jpoison]

45 minutes ago, Netduma Jack said:

Hi, welcome to the forum! Packet Loss isn't usually caused by routers, rather it tends to be upstream equipment like your Ethernet cables, Modem or line to blame.

- Are you testing on a wired connection?
- How are you testing? Is this on Pingplotter?
- Remove the XR700 from your setup. Do you still get Packet Loss?

Yes, wired connection.(using belkin CAT 6A cables)

I got the XR700 yesterday, update to the lastet firmware and did factory reset, i have a ONT from my ISP conected directly to the router so i had to use VLAN, setup QoS correctly.

Start to playing PUBG and in all games that i played i had between 5 to 7 pkt loss spikes, it went to 125% on upload and 2% on download for about 2 seconds.

Removing the XR700 i had no pkt loss with my old router(Asus RT-AC88U)

[Netduma Jack]

2 hours ago, jpoison said:

Yes, wired connection.(using belkin CAT 6A cables)

I got the XR700 yesterday, update to the lastet firmware and did factory reset, i have a ONT from my ISP conected directly to the router so i had to use VLAN, setup QoS correctly.

Start to playing PUBG and in all games that i played i had between 5 to 7 pkt loss spikes, it went to 125% on upload and 2% on download for about 2 seconds.

Removing the XR700 i had no pkt loss with my old router(Asus RT-AC88U)

Alright - the most important thing now is how you're running your tests. Secondary routers (any) can introduce complexity which makes testing your connection less accurate under many conditions. The only way you're going to get a truly accurate reading is to use Pingplotter on a wired PC. If you're seeing packet loss in the game, I guarantee it's not actually happening - and Pingplotter will prove for certain whether it is or not.

[jpoison]

9 minutes ago, Netduma Jack said:

Alright - the most important thing now is how you're running your tests. Secondary routers (any) can introduce complexity which makes testing your connection less accurate under many conditions. The only way you're going to get a truly accurate reading is to use Pingplotter on a wired PC. If you're seeing packet loss in the game, I guarantee it's not actually happening - and Pingplotter will prove for certain whether it is or not.

I will test, but the pkt loss in the game is really happening because i get "teleported" when the spikes begins.

[Netduma Admin]

You'll often see Packet Loss on PUBG servers and it's nothing to do with your connection / router. Easiest way to check is run a Ping Plotter test in an idle connection and see if you suffer from any packet loss. Follow this guide please: http://forum.netduma.com/topic/23881-ping-plotter-quick-guide/

[jpoison]

15 hours ago, Netduma Admin said:

You'll often see Packet Loss on PUBG servers and it's nothing to do with your connection / router. Easiest way to check is run a Ping Plotter test in an idle connection and see if you suffer from any packet loss. Follow this guide please: http://forum.netduma.com/topic/23881-ping-plotter-quick-guide/

I didnt tested pingplotter yet, but don't blame the game, i know that PUBG sometimes have pkt loss issues but not like that, plus when i use my old router those spikes never happen ever, i know its the router problem and you dont want to believe me neither help, so i maybe will send the router back for full refund.

[Netduma Admin]

2 hours ago, jpoison said:

I didnt tested pingplotter yet, but don't blame the game, i know that PUBG sometimes have pkt loss issues but not like that, plus when i use my old router those spikes never happen ever, i know its the router problem and you dont want to believe me neither help, so i maybe will send the router back for full refund.

Could you run a PingPlotter test please as this will confirm for sure. I'm just talking from my experience, once they added the orange connection icons you would often see Packet Loss from the servers appearing. 

Edit: I think one reason they added this was to help them improve their servers. They added it when they moved away from using AWS servers

[jpoison]

1 hour ago, Netduma Admin said:

Could you run a PingPlotter test please as this will confirm for sure. I'm just talking from my experience, once they added the orange connection icons you would often see Packet Loss from the servers appearing. 

 Edit: I think one reason they added this was to help them improve their servers. They added it when they moved away from using AWS servers

what to you mean "orange connection icons" ?

[Netduma Fraser]

I think admin is talking about in game it will show you orange connection icons when there were some connection issues.

[xlr8r]

slightly off topic but since the update on the Xr500....

i ran a ping plotter test on my mates Pc to test the Xr500 on 3 x Fifa (EA) servers and noticed that regardless of what qos settings were used, i got the same spikes on all 3 graphs ... ? Almost as if qos was doing nothing ....?

Tested with xbox one and ps4 being given traffic priority and tried all 3 qos settings on, off and when high priority traffic detected. and 70 / 70

EA servers tested; easo.ea.com , utas.fut.ea.com and utas.s2.fut.ea.com. Streamed 4k and downloaded a huge torrent while testing.

I watched the graph for an hour and graph showed absolutely no difference with any qos settings used ? (same 7/8 spikes throughout the hour)

very strange....? maybe i did something wrong while testing.....

[jpoison]

Well, i think i solve this problem(at least on PUBG), on QoS traffic prioritization instead of using basic Unreal Engine(PUBG, Fortnite etc..) i choose advanced and set source and destination port 7000 to 7999 UDP, played 3 games with no pkt loss spikes, now lets see what happen in the next few days...

[Netduma Jack]

4 hours ago, jpoison said:

Well, i think i solve this problem(at least on PUBG), on QoS traffic prioritization instead of using basic Unreal Engine(PUBG, Fortnite etc..) i choose advanced and set source and destination port 7000 to 7999 UDP, played 3 games with no pkt loss spikes, now lets see what happen in the next few days...

Sounds interesting, lets see what happens over the next few days. Just bear in mind that the accurate way to test for packet loss is through Pingplotter, so I'd recommend running one of those tests at some point.

5 hours ago, xlr8r said:

slightly off topic but since the update on the Xr500....

i ran a ping plotter test on my mates Pc to test the Xr500 on 3 x Fifa (EA) servers and noticed that regardless of what qos settings were used, i got the same spikes on all 3 graphs ... ? Almost as if qos was doing nothing ....?

Tested with xbox one and ps4 being given traffic priority and tried all 3 qos settings on, off and when high priority traffic detected. and 70 / 70

EA servers tested; easo.ea.com , utas.fut.ea.com and utas.s2.fut.ea.com. Streamed 4k and downloaded a huge torrent while testing.

I watched the graph for an hour and graph showed absolutely no difference with any qos settings used ? (same 7/8 spikes throughout the hour)

very strange....? maybe i did something wrong while testing.....

It's never advisable to use game servers as your target URL. Since you're using those servers, the spikes you'll be seeing are highly likely to be related to the target itself which is why you'll see similar results across the board.

Use Twitter.com to run these tests and post screenshots here if you can

[xlr8r]

19 hours ago, Netduma Jack said:

Sounds interesting, lets see what happens over the next few days. Just bear in mind that the accurate way to test for packet loss is through Pingplotter, so I'd recommend running one of those tests at some point.

It's never advisable to use game servers as your target URL. Since you're using those servers, the spikes you'll be seeing are highly likely to be related to the target itself which is why you'll see similar results across the board.

Use Twitter.com to run these tests and post screenshots here if you can

why not use EA game servers for the test...?  given that this was the initial purpose of buying this router...  to eliminate lag spikes especially when on gaming servers  ...?

here is the pingplotter graph for twitter.com (tested while streaming 2 x 4K Youtube videos)

one big spike AFTER it was set to always on...? but after that it certainly does look a little more stable.

can u explain a bit more in detail why i shouldn't be testing EA's game servers so i can get my head around it,.... thanks

 

[Netduma Jack]

3 hours ago, xlr8r said:

why not use EA game servers for the test...?  given that this was the initial purpose of buying this router...  to eliminate lag spikes especially when on gaming servers  ...?

here is the pingplotter graph for twitter.com (tested while streaming 2 x 4K Youtube videos)

one big spike AFTER it was set to always on...? but after that it certainly does look a little more stable.

can u explain a bit more in detail why i shouldn't be testing EA's game servers so i can get my head around it,.... thanks

 

That test looks really positive to me, awesome stuff. With Pingplotter, accuracy is the most important element of the test. Game servers fluctuate a lot on their own - you can have good or bad servers, and they're generally unreliable at the best of times (you can probably tell when you're playing the game!). With a high quality and constantly active server like the Twitter server you can make sure none of the spikes / jitter showing on your graph is actually an issue with the server rather than your connection.

I hope that makes sense - I can't go too technical since I don't know enough about every factor, but there's very good reasons to test your line to a reliable server. Your objective is to make sure YOUR line is stable, not the server itself, so picking a game server doesn't do you any favours.

Also that massive spike is probably just QoS switching modes - could be graphical (i.e a false reading, the spike didn't actually happen) or it might be real, but it'll only happen when you switch QoS modes

[xlr8r]

15 minutes ago, Netduma Jack said:

That test looks really positive to me, awesome stuff. With Pingplotter, accuracy is the most important element of the test. Game servers fluctuate a lot on their own - you can have good or bad servers, and they're generally unreliable at the best of times (you can probably tell when you're playing the game!). With a high quality and constantly active server like the Twitter server you can make sure none of the spikes / jitter showing on your graph is actually an issue with the server rather than your connection.

I hope that makes sense - I can't go too technical since I don't know enough about every factor, but there's very good reasons to test your line to a reliable server. Your objective is to make sure YOUR line is stable, not the server itself, so picking a game server doesn't do you any favours.

Also that massive spike is probably just QoS switching modes - could be graphical (i.e a false reading, the spike didn't actually happen) or it might be real, but it'll only happen when you switch QoS modes

cheers

not sure what that big spike was because it happened just over 1 minute after switching Qos to always on... didnt happen exactly at the same time.

[Netduma Jack]

13 minutes ago, xlr8r said:

cheers

not sure what that big spike was because it happened just over 1 minute after switching Qos to always on... didnt happen exactly at the same time.

Ah okay. Not sure either but it's a one off. You could always try testing again - I doubt it'll keep happening