Spiral Knights - reimburse players CE lost through server restarts.

reimburse players CE lost through server restarts.

8 replies [Last post]

Thu, 08/11/2011 - 04:51

smetchetch1

Legacy Username

it seems to me that since the pvp was added there had been a restart of the server every day.
it says regularly scheduled maintenance, and frankly it sucks...
every day since this started i have wasted ce to reach certain depths (i.e Basil), only to see the server message saying there will be a restart (AKA get kicked to haven) once im half way there.

amongst many other issues i personally have with SK, this is one of the biggest reasons i have to stop playing all together, i have spent more rl money on this game in 6 weeks than i would have playing subscription based games, i do like SK and this is not a hate thread etc.
but there are many sly things that make u use ce.
i bet most of u can think of a time where lag/camera angles/hitbox (lol) etc issues have killed u...

all those other things aside, some form of reimbursement for energy lost (wasted) when servers restart would be a fitting imo.
i completely DO NOT expect this to happen, i am voicing a concern i have as a paying customer
(although i will not spend more rl money on this game until some major issues are fxed and there is more content added)

Thu, 08/11/2011 - 12:21

Facejuances

dont hate me for this, but its true

You dind't waste that ce, you got some crowns/heat/materials for it... unless the server was restarted the second you used the elevator, wich is unlikely since restarts have a short warning.

Thu, 08/11/2011 - 15:11

Starlinvf

They give reminders at 45,

They give reminders at 45, 30, and every 5 minutes after that before the restart, and even gives you the exact time in the MOTD when you login prior to a reboot. You were well informed it was going to happen, and given ample opportunity to get to a safe point (ie elevator) prior to the servers going down. If your on the start of the run, and you see that message, any reasonable person would finish up the level and head up so they don't lose any progress. You were informed, chose to ignore it, well aware of the "risks" involved with your decision.

Its the video game equivalent of a getting a tornado warning an hour ahead of time, but you decide to keep driving into the danger zone because you wanted a smoothie. You know its gonna wreck s*** and when its gonna happen, and yet you went there anyway. On top of that its regularly scheduled, and happens the same time every night. So that means every night, you go for a smoothie, and this tornado comes out of "no where", wrecks everything and ruins your good time..... I just can't sympathize with that kind of idiocy.

And I'm not even gonna go into you spending too much money... If your old enough to buy CE, you should be responsible enough to manage your own spending.

Thu, 08/11/2011 - 22:38

werfjig

Legacy Username

I disagree. I've never gone

I disagree. I've never gone from Basil to the end of a Vanaduke run in 45 minutes, and I can recount at least 3 times I've been in the position where I'd have to cut my losses and restart my run after the reboot. I'm a Scrooge and only do runs using mist energy, so restarting the run would cut into my CE funds which I reserve exclusively for crafting.

I haven't resorted to begging the GMs to escort me to Basil Even if they did offer that service, it wouldn't suffice as a long term solution.

The biggest problem I have is *when* they take the server down. I'm sure they have the means to monitor how many players are currently on spiral knights, and I'm equally confident that their server reboots do *not* take place during time intervals of minimal player traffic (7PM PST? Really?)

Or you can just have regular, scheduled maintenance. The issue here is not to reimburse players, but to not give them an excuse to ask for it.

Fri, 08/12/2011 - 00:33

Wotwotwotwotwotwot

They do have regular server

They do have regular server reboots at 5 AM PST, which is in large part a time of minimal traffic and is done for regular maintenance. Even then, they give you an hour's notice before they reset it and do it at 15 minute intervals, then at the last 15 minutes, they do it every five minutes, then right before the server reboots (2 minute warning).

However, when they do bugfix patches, they could probably implement them better I suppose.

Sun, 08/14/2011 - 01:42

Jeffreysgf

I wasted it, I just beat the

I wasted it, I just beat the JK after reviving, but couldn't get the token. That is a waste.
There is a much better question to be asked, why do they need to reboot the server? It is on the Amazon cloud, that wouldn't be reset. Individual machines might be, but not the entire server. Furthermore, what actually needs to happen? Why can't they do it while the server is running? It doesn't have anything to do with how things are stored, as that would require a patch, which didn't happen. They should just be able to pause things, such as drops, or monster spawns, until after whatever it is they are doing.
Furthermore, this was the first time it ever happened to me, and I only got a 45 minute warning. If I get a 2 hour warning, I won't bother doing a run, but I felt I might be able to make it in that 45 minutes (and if I had another minute, I would have).

Sun, 08/14/2011 - 04:16

Shidara

Legacy Username

What we need...

... is a way to continue where we left off. Instead of getting kicked back to Haven, we would restart where we left off, with monsters present and all.

However, I do faintly remember reading that there was an issue that rendered this impossible. Or perhaps I was mistaken.
I hope for the latter.

Sun, 08/14/2011 - 06:44

Starlinvf

@jeffreysgf: While it is

@jeffreysgf: While it is possible to create a system that can patch code while running, building a process that can do that might not even be possible except in a few old/custom programming languages. But your kind of confusing things... the actual server VM instance is probably still online, but the SK server process (which runs the code) needs to be shutdown, the executable or other resources with old code replaced, and launched again to start using the new code. Its just infinitely easier to shut down, push the update, then restart the server on a clean memory slate. Any soft variables that they want to adjust on the fly would already be implemented that way (like the AFK detection in Blastnetwork).

They could theoretically record the state of all the clockwork instances and rebuild them, but it would require a lot of storage space, would take significantly longer to bring up, and has room for potential bugs in either the save or restore procedures. Then theres the key issue with client connections. 0 Clients connected is the main condition that lets an instance tear itself down, so you'd also need to allow them to live for about 5 minutes to allow people back in... which will compound with the load generated by people rushing to log back in.

Sun, 09/25/2011 - 04:57

Ufana

I don't care about lost CE.

I don't care about lost CE. I'd love to finish a run for once, though.

Search

reimburse players CE lost through server restarts.