People who understand the server side

Thebgj

When the game goes down like this (usually on day 1 of new EQ). Probably means lots and lots of people are trying to play and stressing something. The game does down for “emergency maintenance”.
Then it works again later.

Does that mean that something was actually changed or “fixed”?

Or does it just mean most people gave up and will log in later in term less stress on something and now works for the rest?

I’m only asking because, if they know it’s normal every new EQ, couldn’t they do something more permanent to prevent it?

Deacon

most of the times these days it's less of the "stress" of people playing and more so something that just went wrong. that could be a server going down etc but it's usually not because too many people were playing. they handle that pretty well these days.

so with the new update i'm sure something just sort of happened and needed to be adjusted or changed.

kikiFurieux

You're a bit hard... The game doesn't come down every new EQ, in fact it's been a while since we've been in such situation. They probably figured out something was not right when it went live and they fixed it by performing some update on the server side, maybe some data in the database or something.
I can't believe they would close the game for a certain amount of time on purpose to avoid the rush. We are talking money here, and it's not in their best interest to do so...

dot_ditto

Thebgj said:

When the game goes down like this (usually on day 1 of new EQ). Probably means lots and lots of people are trying to play and stressing something. The game does down for “emergency maintenance”.
Then it works again later.

Does that mean that something was actually changed or “fixed”?

Or does it just mean most people gave up and will log in later in term less stress on something and now works for the rest?

I’m only asking because, if they know it’s normal every new EQ, couldn’t they do something more permanent to prevent it?

I don't know exactly how they have their stuff setup, however, something causing a server to "go down" .. could be a number of things:

the high usage could be causing a variety of issues locally (to the server) .. anything from memory usage, leakage, to disk space usage, to even hardware issues, such as memory, disks, cables, routers, switches .. anything .. really.

Once that happens, any of those various reasons can easily cause the server/system to "crash" .. as it can't easily or safely recover under the situation.

If it's hardware, something like a bad disk, cable, etc ... they could easily swap in a good one, replace, it whatever .. if they have 1 on hand, might only take a few minutes if somebody is on hand and aware of it. Otherwise it could take hours, days or weeks, depending on the exact problem.

If it's just space usage, or memory, or something like that .. it's likely just a "hiccup" in the software or such . (kind of like your home Windows machine getting "punchy" until you reboot it and then everything is fine).
In this case, a simple reboot usually solves the issues.

Based on the timeframes of these outages ... I'm tempted to guess it's closer to the latter .. that is, just the server leaking memory, or cache not working optimally, and things start into a "death spiral" .. (ie things run slower, so take more memory, more memory used, run slower .. running slower, more memory .. etc etc ..)
So whatever, they're probably just rebooting everything, and within a few minutes to an hour, things are backup and going fine.

As I said, however, I'm not privy to what they got going, I'm only speaking as an IT guy myself .. and just speaking at a high level so it's easy to understand to those who aren't

Ktashak

guys you are complicating the things.. i am 99% sure they just restarted the server .. Temp files cleared ,cache cleared, database got a new breeze with emptying temp tables and deadlocked transactions..

GinjabredMonsta

First thing they teach us: "have you tried turning it off and on?"

zuffy

msconfig

Unknown

This content has been removed.

Thebgj

Wow, thank you for all of your responses. I’m sorry if I
T came off as being too hard. Words on the internet (like when texting) can come off anyway the reader wants.

I really was just curious to know what happens behinds the scenes.

@DeaconVelvet
@kikiFurieux
@dot_ditto
@Ktashak
@GinjabredMonsta
@zuffy
@GluteusMaximus

Deacon

no worries man ... you didn't come off as hard at all to me.

DNA3000

Thebgj said:

When the game goes down like this (usually on day 1 of new EQ). Probably means lots and lots of people are trying to play and stressing something. The game does down for “emergency maintenance”.
Then it works again later.

Does that mean that something was actually changed or “fixed”?

Or does it just mean most people gave up and will log in later in term less stress on something and now works for the rest?

I’m only asking because, if they know it’s normal every new EQ, couldn’t they do something more permanent to prevent it?

Generally the former. The game is developed in a dev environment and then the updates are sent to the live environment; the developers are generally not allowed to directly access the live environment except under certain very restricted situations (for a variety of reasons). So when you make an update to a game like this, the update is created in a completely different set of systems from the live game, and then all those updates have to be carefully packaged up and sent to the live environment, very much like Microsoft makes a patch for Windows and then sends it to you to load on your computer. Between the dev and the live environment mistakes or omissions can happen. Sometimes the live environment needs something the dev environment didn't, and so the developers didn't make that thing in the update. Sometimes the update package accidentally fails to include something the developers did make. And sometimes things work in the dev environment but don't work in the live environment at all, and this creates an unforeseen problem.

For example, suppose there's an ability in the game that uses a game engine function, and that function is a little slow. The devs decide to improve it by adding a new game engine function that is more efficient, and then start changing the way those abilities work to use the new function. But there's a subtle flaw in that new engine function that makes it work fine in the dev environment but it doesn't scale in the live game with thousands of players. So when that update goes live and more than a few dozen players start banging on it, that function starts to create performance problems everywhere. This causes outages and disconnects, and the devs start looking at logs and realize where the problem is, so they bring the game down for emergency maintenance and revert those abilities to the older implementation to avoid the problem while they go back to the drawing board to figure out what went wrong with the new engine function.

I'm not saying something like that happens often (or at all), but that's probably not entirely dissimilar to the kinds of problems that can generate an emergency maintenance soon after a new update goes live.

Thebgj

Thanks @DNA3000




Me