CCP_Explorer

All production hardware is transferred to use for test servers for many years.

Probably never.

No.

(Post must be 5 characters, so here are more characters…)

The deployment request was sent by the team at 15:31 this afternoon (today) – how should I have posted that yesterday?

Our network partners are not going to investigate ice belt spawns. Our DBAs are not going to investigate ice belt spawns. Their work is also not going to be put on hold because some other team is investigating ice belt spawns. One mining fix is being deployed tomorrow and I know that team has in general been actively investigating spawn reports.

I hope you enjoyed it.

Official downtime each day is 15 minutes. We only advertise extensions if we expect to exceed that timeframe. On the topic of changes: https://twitter.com/CCP_Explorer/status/1503118118322909197.

Yes; we recently changed how markets are primed; have you noticed?

We prime 1000 solarsystems right after startup. Players who are eagerly waiting to log in, will be logging in as that priming is happening. Then the rest of the systems are primed when the first pilot jumps into them.

The contract system takes a while to prime after startup...

I see that people have had fun in this thread.

I was not amused. The issue encountered today had nothing to do with today’s experiment. Instead we got hit for the 8th time in 11 months with a mysterious issue that Microsoft hasn’t been able to resolve so far. Hopefully the logs this time show them something new – at least the symptom and error was slightly different than before after the latest round of change recommended.

What we were able to do of the experiment did provide information.

Please send a technical support ticket to CCP’s Customer Support. Please include the details asked for in the first post: 20220225 - Connection Issues (UK).

Another view on the overall incident:

image1199×392 48 KB

Status in the last 6 days, closing this incident:

image1140×535 60.3 KB

See 7 March 2022 - Connection Issues (Mexico & South America) - #5 by CCP_Explorer, closing this incident.

Closing this incident.

Anyone?

Our network partners Advania and Cloudflare have been digging into this today and have found network traffic possibly being dropped somewhere inside Level 3’s network, affecting network connections in Santiago CL, Lima PE, Mexico City MX. This would match the symptoms we are seeing. Changes have been made to try and circumvent this possible fault.

Can you all please try again to connect to TQ (without a VPN) and let us know how it goes. Also, did your trace routes change?

@Adelrad and I have been discussing this in his thread Connection to Server was closed, unable to connect on port 26000 but we decided today to start an official thread.

At the same time our internal network partner Advania opened a supp...

We’ve had a few reports recently about this particular issue, but only from EVE players in Latin America – mostly Mexico and Chile. Based on metrics (particularly daily active users from Chile) then this started on 24 February.

The errors are puzzling. Based on the logs we have seen so far then the connection to the Server doesn’t appear to be blocked and it isn’t terminated by the Server or any network equipment. Rather there is this in the logs Invalid answer from server GetServerStatus when the Client tries to authenticate to the Server.

We suspect that somewhere there is possibly some network traffic inspection (filtering or proxy; possibly there is blocking, either a direct block or something that times out) since we are seeing CORS header errors in the Launcher as well, indicating that the headers have been stripped by some middle layer, but primarily that mangled reply to the Client that the Client is unable to decipher. There ...

Even after the fixes before the weekend then the graph was a bit noisy. After the disconnects this morning it is actually a bit more quiet, but then again this is Monday morning and so less people from the UK are online.

image1142×544 65.6 KB

Virgin Media has responded to the email I sent this morning confirming that they received it.

I only find the network diagnostic in one of the ticket and the trace route in the other one appears to be cut off. Can you resend the full trace route into either of the tickets and let us know here, please?