Ok well I’m really glad it’s not a hardware issue, but a connection issue might suck even more… as most of that is out of my control.
I am using a wired connection, and I did reset both the router and the modem. I updated the firmware of the router, and had my isp also update the firmware of the modem.
This unfortunately did not resolve the problem. To add to the mystery… when I log back into the game, after closing it via task manager, the load screen will take much longer than normal. I can then load any other character I have, but if I load the one who I was playing at the time of the problem I will get a “world server” error message. And when I do get into the game I am almost always teleported back to where my hearth stone is set.
I got a copy of that winMTR thing but I keep getting a “can’t post links” message here.
The WinMTR results look good, so no indication of where exactly the partial disconnect is occurring.
Something you could potentially try would be to enable QoS (Quality of Service) in your router’s settings, if your router supports QoS. That can sometimes help with connection issues like this.
Or if you can try connecting through a different connection, such as a hotspot or VPN, that would help identify if it’s an issue happening along the route somewhere between you and the server.
Ok I have been using a VPN with my endpoint in chicago, because apparently my server is located in chicago, and I have not had a problem since.
I assume this indicates that the problem is in the route. Any clue on how I can identify where the problem lies? wan/lan? if it’s lan side obviously it’s something I can change, and if it’s not then I need my ISP to change something somewheres, and they might because we have a business account and they seam to really support business customers more than the residential side.
I’m on a USG Pro 4 and all USG wired switches. Considering that your (presumably commercial endpoint) VPN Solution fixed it, do you have deep packet inspection enabled?
It looks like they own 137.221.96.0/22. I’d like to get a blue’s confirmation before I WL it. I also can’t help but wonder what they are sending that is tripping trojan traffic signatures. I’d like to wireshark it, but the error is so rare, from a packet perspective that I don’t think i’d ever capture it.
Concerningly, this time a different network range in a different part of the world. Unless blizzard decides to take action, we are going to be left with the choice of disabling IPS or suppressing these IP by IP.
I have been attempting to catch it but have not seen the issue yet tonight. As in I have not dc’d yet tonight… still trying to catch it. Hopefully we can both confirm it.
Knoest is probably tired of getting update notices for this thread, but I wanted to add a comment just incase we get a blue reading this.
While I can only be certain that this is happening on Ubiquity hardware, it is reasonable to infer that this will be happening on any high end home networking equipment and virtually all corporate equipment that support DPI and IPS signature detection. In short, this may be impacting a large number of users.
ive seen DPI screw up perfectly clean SMTP traffic because it thought an image in the footer (that had been in there for over a year) was now malicious due to signature matching - giant pain in the you know where because the rule was set to silently drop (which is ultra stupid in a commercial environment)
It was reported in the ubiquity thread I linked. In looking into it further it seems to me that blizz might be initiating a connection remotely to ‘push’ certain data, triggered from your active connection. If I’m not mistaken, that is bad practice and not too shocking it matches a malicious signature – just like any RAT or reverse shell. I’m not too sure what Ubiquity or other vendors could do other than exclude the wow protocol. It also seems like blizz is aware of the problem, but is blaming firewalls generally ("transfer aborted: instance not found” in Shadowlands - #4 by Illìdàn-barthilas)