CCP_Explorer

CCP_Explorer



26 Feb

Comment

I trust that Virgin Media’s logs show more than our logs, and that they find out what happened and fix the issue quickly.

Comment

This morning there were UK-based disconnect events at 09:05, 10:10, 12:45 and then a minor one at 13:30.

Comment

Please submit a ticket to Virgin Media, referencing their ticket F009669229 on this issue.

Comment

That’s both very interesting and weird. You might want to mention this in your ticket with Virgin Media.

Comment

Virgin Media UK has evidently raised an internal ticket F009669229 on this issue. If you reported this issue to Virgin Media UK, please update your tickets with a reference to that ticket. We are still trying to figure out how to get our private contact details into that ticket for direct collaboration on resolving this issue.

EDIT: And now our details are in that ticket; now we wait, standing by to provide information or direct help to Virgin Media to resolve this issue.

Comment

I don’t want to use the word “blame”, but Virgin Media’s senior network admins are the prime position to diagnose and fix this issue, which looks like some sort of a peering issue of some kind with Cloudflare.

Comment

That indeed shows that your local home network is not the issue and also that from Cloudflare* to Tranquility then there are no issues either. So the issue is somewhere in Virgin Media’s network, in between your home and Cloudflare. If you are in the UK and VPN-ing to another place in the UK, you are simply redirecting the traffic slightly to bypass whatever fault or issue there is. I’m very confident that senior network administrators at Virgin Media will be able to resolve this, the issue at hand is simply getting their attention.

*Tranquility is fronted by Cloudflare; your connection to TQ should normally go directly to the nearest Cloudflare...

Read more

25 Feb

Comment

A graph showing how this is only affecting players in the UK and based on reports only Virgin Media customers. We here at CCP can help Virgin Media resolve this issue but we need your help, EVE players, to bring this to their attention.

image1591×632 146 KB

This started at around 22 o’clock UTC on 23 February.

Longer-term view (only showing the UK):

... Read more
Comment

Based on the data we have received so far then only Virgin Media would be able to resolve this (potentially in collaboration with Telia and Cloudflare).

Comment

This is not the first time something like this has happened. Explained in this thread here ...

Read more
Comment

I think they are just showing that the IPs belong to Telia, which is a Swedish company. I think the last hops are in Amsterdam based on what other Geo IP services are telling me and also on the ‘adm’ in the name. My traceroute to that last hop before TQ goes through Lo...

Read more
Comment

Also sent to Virgin Media?

Comment

A common factor in all problematic traceroutes we have seen so far is that these are the last hops:

Read more
Comment

Did you contact your ISP as well?

Comment

Did you contact your ISP as well with that same information?

Comment

Virgin Media customers: Please see this thread 20220225 - Connection Issues (UK) and this tweet ...

Read more
Comment

A number of players have reported that using VPN resolves the connection issues. This is indicative of (changed) QoS settings at your ISP that negatively impact the direct connectivity to Tranquility or BGP routing issues or cross-connect/peering issues, since a VPN connection will send the network traffic through a different route or to a different Cloudflare datacenter.

If you have successfully resolved the connection issues using VPN, then we would want a traceroute both with and without the VPN active, to be able to see the different network paths.

Comment

Virgin Media customers: Please see this thread 20220225 - Connection Issues (UK) and this tweet ...

Read more
Comment

Please note that the traceroute we are looking for is to TQ; on Windows this can be done from a command line like this:

tracert tranquility.servers.eveonline.com

We need this information to try and figure out what Cloudflare datacenter the connections are going through.


02 Feb

Comment

We are very literally moving two main databases to the current passive nodes; since the passive nodes have been fully updated and we need to update the active nodes. We are using the quiet time in between the Interstellar Convergence event and the February release to do this.