Frontier Server Down

[Myrm]

Just been kicked out of Mobius server mid-jump. Now I can't log back in. Can't even get back into Open or Solo

[TorTorden]

Patch ?

God I hope they have a patch coming.

[Myrm]

Must have been a glitch. I'm back in now.

[Myrm]

Patch ?

God I hope they have a patch coming.

A patch? What do they need a patch for? They only released the update yesterday.

[TorTorden]

Galaxy map settings bug out for a lot of players.

Bi weaves are no longer sold ANYWHERE.

Barnacles supposed to drop meta alloys don't.

And that's just three nearly gameplay breaking bugs that I personally know of.

And of course the usual pile of broken missions .

[Schmobius]

They said they are making meta alloys available at a market in Maia, and I think the timing was said to be by about an hour from now, so that could have been it.

[thebs]

Galaxy map settings bug out for a lot of players.

Yeah, I was having a number of issues at times.

Bi weaves are no longer sold ANYWHERE.

Ahhh, now that I noticed after 4 stations.

Barnacles supposed to drop meta alloys don't.

Yeah, I resorted to 'Dukes of Hazzard' style gameplay on Merope 5 C last night (jumping over the Barnacle).

Side Note: I thought they might have just de-activated Merope 5 C for a reason. I.e., well known location in all instances. Or they needed time to respawn (although I waited for hours).

And that's just three nearly gameplay breaking bugs that I personally know of.

All of which I ran into as well over 3 hours of gameplay.

And of course the usual pile of broken missions.

I haven't done any yet, so I guess I'll go back to mining ... for now, at least if the servers are up tonight.

And if not, will be editing last night's 'fun' with a new sound soundtrack ...
https://www.youtube.com/watch?v=OTBy5nNhS24

[thebs]

Patch ? God I hope they have a patch coming.

A patch? What do they need a patch for? They only released the update yesterday.

That's the problem with Beta servers and switchovers. They don't always get the full diff and/or there are unforeseen integration issues, especially when the rules and run-time are utterly different.

A better solution might be to introduce a 3rd set of servers and opt-in ...
- Beta Test (weeks upon weeks, multi-stage)
- Acceptance Test (only a few days)
- Production

► Show Spoiler

So instead of a full cut-over from Beta to Production, where they shut down Beta and everyone is now on Production -- which is not running the same codebase as the new Production -- they have an Acceptance set that matches the new Production instance, verbatim. I.e., full production instances to a subset of opt-in users, with the same codebase that matches production.

The 2.1 Beta didn't have all the logic of Production. Acceptance does. The idea is that you have people opt-in to acceptance test, and there is no guarantee of persistence, if they have to "yank the plug" if they discover any issues. They would have caught all three (3) of these issues, easily and handily, with the subset of users.

Even better ... they would have only had to shutdown the Acceptance Test servers, and kick users back to Production (or Beta), if there was an issue.

In fact, most Acceptance Testing is often done on a Production 1 and Production 2 rotation. I.e., the Acceptance Test servers are actually the next Production servers. E.g., in many environments, it's not a cut-over, but a rotation. X (small set) Acceptance Test servers, to Y (large set) Production servers, then Y-Z=X (existing Production) and X+Z=Y (new Production, formerly Acceptance), where Z (large subset) is the servers that are taken off-line to get the new codebase for the actual changeover without having to take down the environment (just reduction in availability), once the Acceptance Test is verified.

Even better ... this allow people on the old Client version to be assigned to the existing Production servers the Frontier, Steam and other propogation and installation of the Client is complete. It, of course, requires Client and Proxy-Server modifications, but it's something they should aim for in version 3 (the next season).

E.g., the biggest issue is the fact that -- as I understand it -- there is still just 1 persistent galaxy when it comes to procedurally generated systems and worlds, including assigning discovery. Correct me if I'm wrong. Unless, of course, progress in the Beta is rolled back ... including discovering systems and worlds. Then it is separate.

Side Note: Progress roll-back -- in case the first Acceptance Test fails -- may or may not be feasible, depending -- but they could "snapshot" (really more like a "fork") a CMDR's state prior to Acceptance if progress would break a return to Production without the updates. That would be part of the Opt-in agreement, especially since only the "Acceptance Universe" is affected. If no issues with the Acceptance Test, then the Progress could be maintained, as the Acceptance Test servers are the new Production Servers, a reward for those extra 24-72 hours for helping with the Acceptance Testing.

Of course, it's easy to be an armchair DevOps guy. I could be totally ignorant of both some serious concurrency and convergence aspects that makes it impossible. But I've done this for software defined networking and storage (SDN/SDS) -- two things that cannot be out-of-sync without data loss -- and while it takes some effort in the design, the payoffs are well worth it.

As always, I prefer not to complain, but to understand ... even offer solutions (again, from ignorance, but I will any way).

[TorTorden]

Some of the meta alloys drop changes are intentional and they want us to find new barnacle sites.
Which I'm more happy about but apparently the new ones are bugged..

[thebs]

Some of the meta alloys drop changes are intentional and they want us to find new barnacle sites.
Which I'm more happy about but apparently the new ones are bugged..

Yeah, Obsidian Ant just published a new video stating as much.