Tree of Savior Forum

Their decision on regional servers only makes sense of we get 1-2 servers for EA

The chart they provide is a bit missleading, as many have already said, because they split up several EU server related territories to make it look like EU is not a big player base. In reality we have 4 big groups: NA, SA, SEA and EU. all around 25%.
So if they think their EA population is only good for 1 or 2 servers. I can understand their decision. Splitting us up into 4 servers would mean little population. But if they are launching 3-4 servers in EA i will be very disapointed and their statement about regional servers was kind of a lie. if we have players to fill 3-4 servers i would be a natural choice to launch a server in each of the 4 regions since each of them houses roughly the same number of players. And since they are using AWS, it would be very easy to do so.

So i myself am very scared of launch day, because their statement made me believe that they are talking about several servers at EA and if it really boils down to 3+ i think IMC will have lost my trust, because they were giving us lies to defend them using the super easy way out and only hosting in US-east.

What do you think?

4 Likes

You’re reading too much into things. And be aware that “worlds” does not Mean “Servers”.

And ive read somewhere that steam sold 150k EA packs already, so a minimum amount of 3 “worlds” is necessary.

I think this is getting absurd.

They are using Amazon AWS servers. Some people believe that they set up an HQ in places where they host servers. They do not have an HQ in the US, yet they host servers here.

For them to host an EU server, it’s as simple as setting it up with the very easy to use platform they are already using, Amazon AWS, and they have servers in the EU region.

So, with it being fairly easy, and not requiring them to set up some office in a different region, they still have decided not to go through with it.

It’s as simple as they just do not have enough data to support it yet. Come launch, if complaints to happen enmasse and the EU community jumps ship, it’ll be red flags for IMC. Personally I don’t see this happening since for some reason a lot of players are relating MOBA like ping to an MMORPG which is very forgiving. An EU player on reddit was genuinely upset with 100 ping, and I just could not understand it.

But again, regional servers are much more likely to be a thing after launch if there’s a severe drop in the EU population, or a mass amounts of complaints like when the first founderpack info came through.

8 Likes

They already have iCBT1+iCBT2+“the famous 25% EU+Ru piechart”. What more data would they need ?
If they lack data, it’s only about the “future international success” of their game. Forecasts that they should already have since you don’t launch a game without making sure that it’ll be worth it…
Or they may already plan the fail of their game, ofc.

Anyway, their only excuse for refusing the EU players their server, is if they only open less than 3-4 of theses…
But I bet they’ll open at least 3 servers :smiley:

That’s probably what they are waiting. Like how they waited for the 3 months EA complaints.

And be aware that “lambs” does not Mean “Players”.
Your “worlds” thing is funny but… are you high ?

Trust me, it’s more complicated than that, I’m also an engineer who works on a multi-regional service provider.

Some of the simple reasons I can tell you are:

  • They do not have enough people to maintain multiple regional servers right now
  • They do not have enough money to hire the people to do the above
  • They are not confident that the game is polished enough to deploy on multiple regional servers, after all, monitoring 1 server group is easier than multiple (bugs, hacks, exploits etc…)
  • They do not believe other regions will bring enough profit to sustain their own regional server right now (due to their pricing scheme, I must say it’s quite expensive in Asia) so US is their first move to test the water

Of course I am not an IMC staff so those are just my speculations

5 Likes

This most likely.

Also because of the backlash that this caused I doubt EU players will be interested anymore so I don’t ever expect a EU server anymore.

Everything is open to question still and we will only know for certain once the game launches, but the two cornerstones were and are these:

  • How high will our ping be
  • How much will it affect gameplay

If one or both of these factors will create a situation in which all/certain classes are less/unplayable, then we have an issue which will escalate really really quickly and spiral down into certain regions simply leaving the game to hell.
IMC’s standpoint of “we need more support to justify a new server” is a paradox because if there will be issues due to their decision of not opening the eu server, then by definition there will be no more support. They would need the server exactly in order to get more support.
Its borderline retarded to imagine that there will be growing support for a service which is flawed.
The reality of the question at this point is that if the game is screwed by high pings, then they have lost half the eu region and everything else east of central europe.

From what i’ve seen this game doesn’t have an ability queue, so that means if you have 100ms ping, expect to have 100ms delay between all skills including instant no cooldown skills. You can imagine how ridiculous it quickly becomes once you start to climb above 100ms

And there we have it.
If its Rangarok’s jumping across the screen with skills firing 0.5-2 seconds later than clicked, then its mental for IMC to expect more support for ToS.
They have it completely backwards:
You can expect more support if you have a flawless service and good playing experience. NOT when you dont give 2 ****s to as much as add an eu server for an international release, saying that you dont see it justified.

When they lose next to all players outside of NA i guess there will be some morals and lessons to be examined.
Too bad that it will be too late by then.

Hah. Seems like you’re getting somewhat spiteful lately when you’re starting to nitpick my autocorrected stuff. You seem to have missed the point of the Staff only adressing any “servers” as “worlds” lately. If you don’t believe me, go check the announcements.

Why do you think they’re doing that? Its possible the “worlds” are all located on the same physical server. Meaning, they may be able to add additional worlds to play on, but not on a different location.

What now, more petty insults?

They’re doing that cause Role Playing. You know, it may come as a surprise to you but you’re not a “Savior” IRL…
Everything is “possible” until you know(=have proof) that it “is” or “isn’t”.

i thought about this too, but if this is their architecture then the game is doomed anyhow.
Even though the are using server virtualization (AWS in this case), no provider can scale one virtual server endless. What this means is that there will be a natural limit to worlds they can create and we all must pray that this is a rather high number. But considering how much power typical MMO servers demand and looking at the biggest virtual machines AWS provides (yes you can get more power for a custom machine, but trust me, prices rise exponential if you move past the predefined tiers) this could really mean grim times at the f2p launch.

IMC looks confident being able to provide more capacities if needed, as in adding “worlds” virtually using AWS. Can’t see a problem here if amazon and steam gives their ok.

You know whats ironic? @lemaitre_dulotus mentioned IMC probably using AWS or other cheap server-farm architectures himself days ago and now tries to ignore the obvious signs, calling them doing “roleplay”. I always wondered why they needed to call the new servers (laima, gabija etc) worlds when everyone expected the term “servers”, until someone else mentioned that they don’t actually add physical servers at all.

After thinking about it, that makes sense. It’s cheaper to only add capacity if needed instead of running already rented servers idle, knowing that IMC goes the cheapest route possible.

Going for that architecture gives them somewhat freedom about the accountmanagement, as in being able to move teams from world A to world B without much effort.

At the same time, they’re kinda bound to the location. Means, to open “worlds” on other physical locations, they need the whole architecture again instead of just one more server on another continent.

As a developing company using years for developing without actually earning a penny, IMC is dependent to investors who only invest money in exchange for proper data - data IMC can’t deliver yet. The only data they’re able to deliver are the EA sales, which are mainly worthless due to many users holding back waiting for news about regional servers. Thats a serious dilemma they finally managed to overcome by promising a transfer to future regional servers, opening the way for European/Sea players to finally buy the EA and giving IMC finally the data they need to get investors providing the money to open these servers.

Quite ironic huh? Guess I need to write that into a topic >,>

Allow me to clear the facts.

Yes. Not that I did check the facts but I believe people saying that they’re using AWS, especially since I’m pretty sure that they’re using Amazon servers.

I said calling servers “worlds” is roleplay.
Which means that I think that Laima is one instance, Gabija is one instance, etc… (An instance is a virtual server, that you can indeed easily grow or shrink depending on your needs.)

That’s still what I call “roleplay”.
Only a company like imc would call cloud servers “worlds”. Nothing is wrong in that, but don’t overthink it.
For the customer (the player), there is no difference between a cloud server and a dedicated server. A cloud server is still a server. It’s only a convenience for the publisher. Like you said, it allows way more flexibility. It has nothing to do with account management though…

We may have opposite opinions but I will not let you call me a weathercock when you don’t understand what I’m meaning ^^

Then I’d advise you to stop being so hostile when you don’t actually have anything hostile to say, besides being hostile!

Or just replying to me for the sake of replying when you don’t have any other intentions as just trying to prove me wrong, actually being unable to do so by admitting it in your own posts between the lines.

Since IMC can’t call it servers as we’re used to and expecting it, they’re calling it worlds. Having a world/section next to another means they are physically in the same network (or close) to provide as less latency as possible. Sourcing them out will cause costs. Money IMC doesn’t have. And thats all.

lets be honst: all this is speculation. We don’t know what their setup really is.
I can just speak from experience how the usual MMO setup would work:

You would have one login server which then is connected to dedicated servers for the different regions. In AWS/cloud/tos vocabulary we would have one instance which is the login server which is then connected to several other instances which are the worlds.
If this is the case then nothing would stop them to host the instances/worlds in different regions via AWS and all connect them to the one login server that can be hosted wherever it suits them best. There is no extra work, costs etc involved, if they set it up properly.

The other possible solution is them using one instance for all: So they have one AWS instance that houses the login server and all the worlds. If they need extra worlds they would increase the virtual hardware of the instance to be able to house more worlds. But this is a very bad structure from a professional point of view.
For once: having the login server on the same machine/instance as the game servers/worlds means that they share the same resources. So whenever their is a maintenance, any other server restart or them just opening the game to everyone, or to a new region that was blocked etc. in general any situation the login servers get hammered by lots of people at once, the worlds will suffer, too. everyone who is already in or was in, will experience lags because of the login server sucking up all the cpu/ram etc of the virtual machine.
second: whenever you create a new world or raise the instances power there is a downtime for the whole service.
third: You cannot scale a virtual instance endlessly and especially not cost efficiently. I talked about this before, but the reality is that AWS can only provide a certain maximum of cores and ram per instance and once that limit is reached you would need to contact them for a custom solution and thats when it gets REALLY expensive.

So from my perspective there are two options: either they have a good server structure and just dont want to give us regional servers even though they could do easily OR they have bad infrastructure and need to fix that first before even thinking about regional servers. Pick your poison.

1 Like

Expect them to have picked the most cost efficient solution fitting their needs as company without an infrastructur and much experience. Sourcing worlds out might be technically possible but they cant sustain them yet and provide support the way they want/need it.

The picture we have right now says the Investors nay’d to expansions. Only if IMC can provide data saying expansion to other regions will be worth it will lead us to get these regional Servers.