Roll20 uses cookies to improve your experience on our site. Cookies enable you to enjoy certain features, social sharing functionality, and tailor message and display ads to your interests on our site and others. They also help us understand how our site is being used. By continuing to use our site, you consent to our use of cookies. Update your cookie preferences .
×
Create a free account

Our GM Can't Connect to the Roll20 Website

1407613812
The Aaron
Roll20 Production Team
API Scripter
Can you check your console log for errors, and check to see if there is a particular host that is timing out on the network tab?
Not terribly skilled at network troubleshooting so I took screenshots of the console log and the network tab - does this help? <a href="http://imgur.com/a/cpQ7w" rel="nofollow">http://imgur.com/a/cpQ7w</a>
1407641944
The Aaron
Roll20 Production Team
API Scripter
Kudaku: I know you said 9/10 of the time it fails, are those from a failed attempt?
1407649562
Karl V.
Pro
Translator
To my fellow Scandinaviian's how are suffering from this: The TOR browser work-around that Kenny posted up-trhread does indeed work. The only downsides I have found are: 1) Its a bit slow. Probably because it routes your traffic through the TOR network 2) It does not support WebRTC, so you will have to use Skype or similar for voice (but most ppl do anyway...)
1407689821

Edited 1407690148
Aaron said: Kudaku: I know you said 9/10 of the time it fails, are those from a failed attempt? Yes they are. Are they any good? Is there anything else I can do to help? Edit: Trying out the TOR browser now, it seems to sidestep the problem completely.
I've reached out to our hosting provider tos ee if they have any insight here. Not a whole lot else we can do unfortunately. It seems pretty much confirmed based on what we're seeing here that this is a network path issue that's only affecting Scandanavian folks. :\
They replied back that they want us to run this diagnostic tool: From the affected location: mtr -rwc 100 &lt;IP ADDRESS OF SERVER&gt; More information on generating and interpreting MTR reports can be found in the Linode Library: <a href="https://www.linode.com/docs/networking/diagnosing" rel="nofollow">https://www.linode.com/docs/networking/diagnosing</a>... You can get to the tool for Windows here: <a href="http://winmtr.net/" rel="nofollow">http://winmtr.net/</a> The IP address of our server is 66.228.53.175 So, after you run the MTR tool, please PM me your own IP address (you can find it by going to Google and just searching for "What is my IP Address" and Google will tell you) so I can provide the other "half" of the report to them. Thanks!
1407769803
The Aaron
Roll20 Production Team
API Scripter
If you're using a Mac, Linode's instructions use mac ports to install. If you have Homebrew installed, you can use ( your input in bold & underline) : &gt; brew install mtr Also, they don't explicitly note this, but on a Mac or Linux , you need to run it either as the root user, &gt; su - root Password: ******** &gt; mtr -rwc 100 66.228.53.175 or with the sudo command: &gt; sudo mtr -rwc 100 66.228.53.175 Password: ******** ( Note: with su or sudo, the Password is the current user's password, not the root password. )
PM sent!
Belgium too! I am on vacay for a week but will run the diagnostic when i get back!
1407857605

Edited 1407857916
Riley D.
Roll20 Team
I've gotten two PMs with logs, thanks! From what we're seeing after consulting with our hosting provider, basically the problem is "on the other side of the ocean" -- the packet loss is being experienced in the connection between your ISP and some of the backbone providers (e.g. Level3). However, I'm not really sure why you would only be experiencing the issue on Roll20 if that's the case, as other sites (such as Netflix) also use them as a provider. For example, here is a log from someone showing around 20% packet loss at the Level3 connection: | ae51.bar1.Oslo1.Level3.net - 0 | 373 | 373 | 7 | 17 | 117 | 11 | | ae-7-9.bar1.Copenhagen2.Level3.net - 12 | 257 | 228 | 22 | 31 | 103 | 26 | | ae-0-10.bar1.Copenhagen1.Level3.net - 20 | 209 | 168 | 19 | 24 | 44 | 27 | | ae-7-7.ebr1.Dusseldorf1.Level3.net - 0 | 372 | 372 | 110 | 117 | 197 | 113 | | ae-21-21.ebr2.Dusseldorf1.Level3.net - 0 | 372 | 372 | 110 | 118 | 189 | 117 | | ae-45-45.ebr1.Amsterdam1.Level3.net - 0 | 372 | 372 | 110 | 117 | 193 | 115 | | ae-23-23.ebr2.Amsterdam1.Level3.net - 0 | 372 | 372 | 110 | 117 | 185 | 114 | | ae-47-47.ebr2.London1.Level3.net - 0 | 372 | 372 | 114 | 121 | 203 | 118 | | ae-43-43.ebr1.NewYork1.Level3.net - 0 | 372 | 372 | 113 | 120 | 206 | 118 | | ae-91-91.csw4.NewYork1.Level3.net - 0 | 372 | 372 | 114 | 120 | 199 | 119 | | ae-4-90.edge3.NewYork1.Level3.net - 0 | 372 | 372 | 114 | 123 | 201 | 116 | | te2-5.bbr01.tl01.nyc01.networklayer.com - 0 | 372 | 372 | 105 | 112 | 186 | 111 | | ae1.bbr01.eq01.chi01.networklayer.com - 0 | 372 | 372 | 135 | 141 | 209 | 137 | | ae20.bbr01.eq01.dal03.networklayer.com - 0 | 372 | 372 | 157 | 164 | 233 | 165 | | po31.dsr02.dllstx3.networklayer.com - 0 | 372 | 372 | 149 | 156 | 231 | 152 | | po32.dsr02.dllstx2.networklayer.com - 0 | 372 | 372 | 159 | 167 | 346 | 169 | | po2.car01.dllstx2.networklayer.com - 0 | 372 | 372 | 157 | 167 | 256 | 166 | | router1-dal.linode.com - 0 | 372 | 372 | 157 | 164 | 236 | 159 | | li295-175.members.linode.com - 0 | 372 | 372 | 149 | 154 | 225 | 151 | |________________________________________________|______|______|______|______|______|______| Here's another log from someone else who didn't go through Level3 showing no packet loss: | kbn-b3-link.telia.net - 1 | 199 | 198 | 9 | 14 | 69 | 10 | | kbn-bb4-link.telia.net - 0 | 203 | 203 | 9 | 23 | 43 | 9 | | nyk-bb2-link.telia.net - 0 | 203 | 203 | 102 | 117 | 204 | 106 | | nyk-b6-link.telia.net - 1 | 199 | 198 | 97 | 105 | 170 | 98 | |xe-0-0-1.bbr02.tl01.nyc01.networklayer.com - 0 | 203 | 203 | 103 | 111 | 163 | 103 | | ae1.bbr01.eq01.chi01.networklayer.com - 0 | 203 | 203 | 123 | 133 | 156 | 124 | | ae20.bbr01.eq01.dal03.networklayer.com - 1 | 199 | 198 | 139 | 146 | 214 | 140 | | po31.dsr01.dllstx3.networklayer.com - 1 | 199 | 198 | 144 | 150 | 203 | 146 | | po31.dsr01.dllstx2.networklayer.com - 0 | 203 | 203 | 144 | 147 | 203 | 145 | | po1.car01.dllstx2.networklayer.com - 2 | 195 | 193 | 146 | 160 | 230 | 147 | | router1-dal.linode.com - 0 | 203 | 203 | 139 | 146 | 200 | 140 | | li295-175.members.linode.com - 0 | 203 | 203 | 145 | 152 | 179 | 146 | |________________________________________________|______|______|______|______|______|______| WinMTR v0.92 GPL V2 by Appnor MSP - Fully Managed Hosting & Cloud Provider The weird thing is that to some degree Internet connections are supposed to be able to "be okay" with *some* level of packet loss..it should just be slower? Another way to look at this: notice how in the end it always goes through "networklayer.com"? That's because that's who our hosting provider uses as their own backbone provider. So the traffic first goes to your ISP, then gets routed through the backbone on their end, then across the ocean, then to the hosting backbone provider, then to our server. It would appear that from Scandanavia there are (at least) two options: going through Level3 or going through telia. It seems telia is doing okay but Level3 is having issues. I'm not sure that there's any way to "force" your traffic to go through one over the other, though...? In addition to that, we haven't changed anything on our end. This problem started probably due to some issues either with your ISP or with Level3 internally. Hopefully they'll also fix it (although I'm surprised it's taken this long to do so)...but that means it's probably out of our hands, too... Honestly, we're way over my head in terms of what to do here (I'm not networking expert by any means). I think if you show your MTR log to your ISP they should be able to help further in tracking down the issue. The hosting provider on our end is basically telling us "this has nothing to do with our end of things", since the packet loss seems to be occurring way closer to Scandanavia than it does closer to us in the USA.
1407860767
The Aaron
Roll20 Production Team
API Scripter
I think the 20% ICMP packet loss on that Level3 router is probably just rate limiting. If it were dropping network traffic, it would be ~20% loss for each hop after it, as that traffic must implicitly pass through that server. Further, it would affect all traffic, not just Roll20. Trace route can really only diagnose network connection issues, not protocol problems. The fact that the reports show very little loss on the final hop means they are not having a problem connecting across the network to that particular server. I agree that the issue is completely outside the control of Roll20/Linode. It seems to be specific to the connection to Roll20, but it is probably really specific to the types of connections to Roll20 or the type of data being exchanged between Roll20 and the client browser. Somethings to consider: Some ISP/Router could be blocking higher ports that are specifically know to be used by certain viruses to communicate, but which also happen to be in the range of some of the secondary services roll20 employes, like Firebase. Some ISP/Router could have a problem with how it handles TCP Packet fragmentation and reassembly, mangling data packets the client depends on and causing it not to load properly. The TOR browser appears to fix this problem because it re-encodes the information and transfers it in a different way. If you could look at the traffic going through the TOR browser, you would see that it is not on the same port ( probably all on 443 ) and not the same size ( probably broken up into many small packets to pass through the encrypted tunnel ), and not following the same routes ( which probably has less of an effect on this data ). Your best bet is to contact your ISP and tell them exactly what problem you're having, and when it started. Try to get escalated to a higher level of support staff as quickly as possible ( ask for 2nd tier or 3rd tier support or whatever the local term is ). Make sure you tell them you are using a web based application with Javascript Sockets on multiple ports, and that the issue appears to only be affecting you and other users on your ISP's network (though really, it is probably affecting everyone in your regional network, for some definition of "region".)
1407861828
Karl V.
Pro
Translator
I am at tier 2 with my ISP atm. I got an SMS to confirm that a technician was working on it a few hours ago, and right now the site is working perfectly. Now clue if he did anything, or if it is just better temporarily.
1407861983
The Aaron
Roll20 Production Team
API Scripter
AWESOME! It would be interesting to see if others in the region are experiencing similar success.
1407863770
Karl V.
Pro
Translator
I am not calling it a success just yet, but right now everything works like normal.
The website seems to be working well now - I'll keep checking every few hours in case it starts acting up again. Thanks for the help so far! :)
1407953703
Karl V.
Pro
Translator
Still working fine. I am willing to call it a win now :)
Kudaku said: The website seems to be working well now - I'll keep checking every few hours in case it starts acting up again. Thanks for the help so far! :) Haven't had any issues for the last 24 hours, looks like whatever it was has been sorted for now. :)
I'm here to report that it seems to be working again as well. I havent really been able to follow this thread for a while now since when the forum actually loads, it stops halfway and only shows a few posts. I'm happy that it finally works though since now we can actually finish the game we once started =P
1408015273
The Aaron
Roll20 Production Team
API Scripter
That's great news!!
I would also like to report that the site is working for me as well. Basically the same as Niklas said over here. I guess i can get rid of the tor browser, it couldn't run the in game things smoothly anyways.