Solved

Terrible Latency from Paris Datacenter Since 25th December

  • 27 December 2020
  • 12 replies
  • 314 views

Since 25th December I’ve been getting terrible latency from my Shadow at Paris datacenter making it virtually unusable. I contacted support and they suggested it is a problem with my network. However, I am pretty sure it is not. Attached you can see a speedtest for the Paris datacenter which has low download, high ping and lots of jitter. I have also run one for the Amsterdam datacenter and it looks normal.

 

I have also attached a traceroute to try and help diagnose the issue. Any help would be much appreciated, it is frustrating not being able to use Shadow whilst I am on holiday.

+++

  • Application type: Mac

  • Shadow Client Version: Official

  • Local device CPU: Dual-Core Intel Core m5

  • Local device GPU: Intel HD Graphics 515 1536 MB

  • Datacenter location: Paris

  • Connection speed: See attatched

    Speed test paris
    Speed test Amsterdam
    Traceroute Paris

     

icon

Best answer by Jim29er 28 December 2020, 16:42

View original

This topic has been closed for comments

12 replies

Userlevel 5
Badge +4

The support folks are essentially correct. It may not be a problem with “your” network per-se, but there are potentially problems somewhere along the path. Shadow (Blade) has no control over which routes ISPs take to get to their systems, and those routes are impacted by things such as congestion and ISP peering decisions. That Amsterdam is “good,” simply means that there are no similar issues along that path.

Also, your looking glass screenshot is a traceroute from the Shadow data center to you, and that path is likely different for the reverse (your path to the Shadow data center). To look at the latter, you need to look at a traceroute from your system to a Paris endpoint (speedtest.shadow.guru [85.190.65.4]).

traceroute speedtest.shadow.guru

 

I have also been experiencing this problem. I opened a ticket with support on the 24th and have been going back and forth with them ruling out all potential issues. 
 

I can also confirm that it does seem to be a problem between the ISP and just the Paris datacenter. 


I have tested the issue across multiple devices with my internet and been able to recreate it.

I have also speed tested on a friends router and everything worked fine. 
 

I am based in London and my ISP is Hyperoptic 1GB. I plugged an ethernet cable into my internet port and directly into my Mac and then tested again on my PC. The results were the same across both, confirming the OP issue.
 

I have attached screenshots below of multiple speedtests.

Now I am trying to establish the best action to take. 

 



 

Interesting, I am also using Hyperoptic but from Manchester and have tried many of the same things as you.

 

It has been marginally better for me today but not back to how it was pre-Christmas. Post here if you discover anything and I’ll do the same.

Interestingly I just tried using a VPN service with a server in France and the issue disappeared. I guess we need to talk to Hyperoptic then. 

Userlevel 5
Badge +4

Do either of you happen to use Nvidia’s GeForce NOW service? I’m curious if you have similar issues with their Paris region (EU Central 4).

When VPNs make these sorts of problems go away (which is not uncommon for GFN as well), it makes me wonder which aspect of the VPN is “masking” the original problem - because the traffic is now tunneled, and/or that it’s using a different route.

El mismo problema aquí, imposible de jugar y sin solución después de reiniciar. 

Hi guys, 

 

Update from my side. I ran a speed test via the Paris data centre this morning and the issue seems to have disappeared. I haven’t changed anything my side. The speeds may seem less in the screenshot but this is because I am using wifi rather than ethernet into the wall. However, everything is the same as the test results I posted yesterday! 

 

I emailed Shadow support late last night and I got a reply asking for traceroutes from ly local PC to Paris data centre and the reverse. They also said they were investigating.


I haven't supplied them with the traceroutes yet but not sure if they have actioned something their end overnight which has resolved the issue. 

 

Magnetesk, did you speak to Hyperoptic at all? I'm stumped as to why the issue seems to have resolved itself. Also, hoping it has fixed it for you? 

 

Jim, I can confirm i don’t use Nvidia’s GeForce NOW service. 

 



 

It has resolved for me also.

No I didn’t get around to contacting Hyperoptic - it was on my list of stuff to do today but it looks like it wont be necessary.

I am also puzzled as to why it has suddenly returned to normal - looking at new traceroutes it looks like it is using different peering nodes now - don’t know if Hyperoptic has changed something in its routing?

 

Glad that it is resolved for you also Shaunfk :grin:

Userlevel 5
Badge +4

Glad to see it’s working for you now...I do suspect Hyperoptic adjusted routing to avoid the problem areas. ISPs are supposed to do that “automatically,” but perhaps it took some time to bubble up.

Userlevel 5
Badge +5

@Magnatesk 

 

The best thing you can do when you have issues like this in the future is to send logs asap! They help shadow immensely!

-Gelgoog 

Badge

El mismo problema aquí, imposible de jugar y sin solución después de reiniciar. 

Bro, arreglaste el problema?

Userlevel 2
Badge

@Magnatesk

 

The best thing you can do when you have issues like this in the future is to send logs asap! They help shadow immensely!

-Gelgoog 

Would be great if that could be done by just pressing a button. A lot is logged already so if they just add things like traceroute and such then clicking the button will make everyones live easier.