Help us test our new Mega-Server on Thursday!

[Ecrir 887]

Bad time for me, else I'd be there. I somewhat expected that a test like this would take place in the weekend, not in the middle of the week, though I have no idea at which time of the day/week Wurm usually has the highest logged in player count.

 

 

This server is big enough to fit all the other pve servers, isn't it? If it works out lag wise, couldn't you even move/copy over the existing servers onto it? You'd get one big server with several islands then (and some people with multiple deeds I guess). There would probably even be some areas left where new landmasses could be created. Does sound like a huge pain to accomplish from the database end.

[rosedragon 2,282]

double post.. sorry 

Edited April 10, 2014 by rosedragon

[rosedragon 2,282]

Time to start maping the server . Also can we test the new tutorial pweassse?

 

one big server.... and what will be with other servers?  why cant join all in one huge map + add more space

If I am not wrong, the old maps are using old codes which make them laggy and not optimized. Might also preparation for new codes like 3d mining o.o ....

[Farmerbob 369]

Can we have a mass drop and take session? 

 

What about Barrels in Raft Cart-to-FSB corn transfers?  Who can Unload 10,000 corn into their FSB first?

[Weponsmith 36]

What about Barrels in Raft Cart-to-FSB corn transfers?  Who can Unload 10,000 corn into their FSB first?

Mabey you missed the part where you can only unload 100 items in a single drag now?

Oh, you would have to actualy be caught up on not-so-current event's, my bad.

[Kamill 61]

Finally living in Europe pays out.

 

To be completely honest, if one of the small Epic servers can't handle ~100 people in local (not even that) without everyone crashing to death, how is this new server gonna handle anything?  I don't really know much about servers and lag, etc., but just seems weird to me....

 

But will def be there, can't wait to see a giant server!

 

I too am curious to see how this works. I can see 200-300 people logging in at the same time, crashing the server(s) instantly for a while.

[Badgirl 115]

Maybe this is how Rolf intends to add flight to the game. With a server that big, you'd need it to travel far distances in a reasonable amount of time.

[Gavin 971]

Dang; test starts at 4am and I've got work at 5 Will the server be up for at least 12 hours or so?

 

-edit- just read; seems kinda pointless running a stress test when everyone is going to be at work on a weekday o.O

Edited April 10, 2014 by Gavin

[Derp 310]

12:00-13:00 CET on a Saturday would be more successful....

Edited April 10, 2014 by Derp

[Ecrir 887]

Maybe this is how Rolf intends to add flight to the game. With a server that big, you'd need it to travel far distances in a reasonable amount of time.

 

I don't think flight would be required, it would only make the world feel a lot smaller than it really is, which would be a shame when you've finally got a big server. Without flight people will probably develop more regional communities, though it is also possible that there will be a lot of water in which case fast boats would probably do the trick already. Personally I'm hoping it will mostly be one big landmass so there's more land based travel and deeding near water doesn't give as huge an advantage as it currently does.

Edited April 10, 2014 by Ecrir

[Sharkin 1,319]

Dang; test starts at 4am and I've got work at 5 Will the server be up for at least 12 hours or so?

 

-edit- just read; seems kinda pointless running a stress test when everyone is going to be at work on a weekday o.O

I can't confirm, but I heard 1 hour o.O

[Ecrir 887]

To be completely honest, if one of the small Epic servers can't handle ~100 people in local (not even that) without everyone crashing to death, how is this new server gonna handle anything?  I don't really know much about servers and lag, etc., but just seems weird to me....

 

But will def be there, can't wait to see a giant server!

 

 

I assume there will be some serious hardware differences between a small Epic server and this giant server. That small server probably doesn't have the resources available for ~100 people in local because it wasn't assigned resources with that in mind. I believe Wild/Indy can also handle such a situation a bit better and Wild has been around for quite some time at it's current size already, 7+ years now? Hardware might have advanced to the point where a 16 times as big server might be viable, especially ram wise.

[Bladedancer 20]

I'll test it because I'm curious, but I'd still like to know the point of this server.

[Theodis 339]

Personally I'm hoping it will mostly be one big landmass so there's more land based travel and deeding near water doesn't give as huge an advantage as it currently does.

Edited April 10, 2014 by Theodis

[Pwent 113]

A few ways to make this server unique and make me move there:

1) Change trader functionality

2) Respawning of uniques, remove drake and scale from dragons.

3) Change back regional veins like they were on Indy, where you had a small chance to find any vein anywhere despite them being regional.

4) A large steppe/desert/tundra (preferably one of each at least) for hunting where no deeds or hunting shacks can be built.

5) Connect it to both Pristine/Release and Freedom cluster.

[RockyBalboa 565]

######

[Jakenheimer 63]

I hope you guys take lots of screenshots for those of us who can't play.

[Devily 623]

A few ways to make this server unique and make me move there:

1) Change trader functionality

2) Respawning of uniques, remove drake and scale from dragons.

3) Change back regional veins like they were on Indy, where you had a small chance to find any vein anywhere despite them being regional.

4) A large steppe/desert/tundra (preferably one of each at least) for hunting where no deeds or hunting shacks can be built.

5) Connect it to both Pristine/Release and Freedom cluster.

 

I agree on everything except the remove drake and scales from dragons. Lets say get less from dragons as u do now and yes.. make them connect to all servers so every one can benefit from the new world.

 

I would love to go on it!

[Gavin 971]

4) A large steppe/desert/tundra (preferably one of each at least) for hunting where no deeds or hunting shacks can be built.

 

This isn't a server you live on for hunting. To put this into perspective;

50,000 animals

268,435,456 tiles

5,368 tiles per animal

 

versus Deli

50,000 animals

4,194,304 tiles

83.88 tiles per animal

 

Statistically it's gonna have 64 times worse hunting than Deli/Exo/Cele etc. This is basically a sandbox server where you don't have to worry about mob's. I somewhat expect that people will be encouraged to move here from the existing servers and eventually the Freedom cluster might just be Chaos (PvP) + Hunting (some tiny like 512 x 512 server with 100,000 animal cap) + Xanadu (sandbox)

[Manny 380]

The animal count is 50k for the test, it's not the maximum amount

[Kagrenac 812]

I assume there will be some serious hardware differences between a small Epic server and this giant server. That small server probably doesn't have the resources available for ~100 people in local because it wasn't assigned resources with that in mind. I believe Wild/Indy can also handle such a situation a bit better and Wild has been around for quite some time at it's current size already, 7+ years now? Hardware might have advanced to the point where a 16 times as big server might be viable, especially ram wise.

The issue is wurms client, not the server.

[Ecrir 887]

I don't think you understand how large the map is going to be. The total number of square kilometers for all the maps currently combined is 1040. This map is going to be 4096 square kilometers. That's nearly 4 times the amount of space as all the current server combined. One landmass would be ridiculous. This can easily house a number of continents larger than indie as well as an archipelago of *hundreds* of affliction sized islands. The ridiculous thing is the animal count of 50k. If we say half the map is water then there will be roughly 24.4 animals per kilometer of land. One square kilometer is 1 million square meters. So yeah... if 1 tile is 4 meters by 4 meters or 16 meters squared that means about 1 animal per 2,560 tiles and that's under the assumption that half the map is water. If like the rest of the servers it's mostly land then the wildlife will be even less dense. Good luck finding a pair of breeding animals. Going to take you weeks or maybe months of searching.

But since the wildlife is so spare this is effectively the peaceful server. Odds are if you don't move around much you'll be able to go weeks or maybe months without seeing an animal that isn't domestic.

I get the feeling when this was being planned someone didn't do the math. To get anything near reasonable animal density the number of animals that need to be on the server will need to be increased by several orders of magnitude and the way servers already buckle under the animal load I just don't see that being possible without disabling animal AI entirely and effectively turning them into animated objects.

 

I know it's that big, which means real mountains might for example finally be possible instead of the small rocks we currently have. Same for real deserts, not something you cross in 5-10 minutes and borders marsh. You could have actual big biomes with such a landmass, resulting in really distinct areas.

 

As for the animal issue, if a reasonable animal count cannot be reached without serious performance issues then indeed the system should be reworked, for example by making it slightly more instanced. For example, say that the server instances animals around the player (in such a way that it isn't visually noticeable, so they spawn before you get within vision range) depending on the lairs/deeds nearby and on how many animals have recently been killed in the area (the more that have recently been killed, the less which will be spawned). They will also be despawned again if there haven't been players nearby for some time (depending on the time chosen, you might still encounter the animal one hour later, but not 2, which wouldn't be strange as it probably just moved somewhere else).

Of course an issue is taming/breeding, but that shouldn't be a problem. For example say if you've put a rope around an animal and are leading it, or have hooked it up to a cart, or it's tamed, then it cannot despawn. Same if the animal is on a deed or if it has been branded. This way you can take an animal with you, put it on your deed, and it won't go away when you aren't around.

In essence it shouldn't change taming/breeding as animals can be made not to despawn, while at the same time making it so that players encounter enough animals out in the wild, but for hunting they do need to keep moving around as hunting in one area would decrease the amount of mobs spawning there. It gives players the illusion of wildlife existing all over, with lairs still determining what spawns where, without actually requiring that 1 animal exists per every, say, 500-1000 tiles (that would normally require 1 million creatures, assuming no water). AI would also be slightly less of an issue if the animal just despawns when nobody is around.

 

That's the direction I would take the animal spawning into for such a huge server. Added advantage is that you'd get rid of stuck animals and animals on steep cliffs, as they'd despawn when nobody is around. Actually simulating every animal like how it's done now seems like such an inefficient usage of server resources.

Edited April 10, 2014 by Ecrir

[Willis 18]

The animal count is 50k for the test, it's not the maximum amount

What was said before.

[Lostone 93]

I think some players eyes glaze over and don't read whats posted.

Such as its a test only 50k animals for now.

Plus you will be spawning in different locations , not everyone in same area.

[Kyrmius 193]

I don't think you understand how large the map is going to be. The total number of square kilometers for all the maps currently combined is 1040. This map is going to be 4096 square kilometers. That's nearly 4 times the amount of space as all the current server combined. One landmass would be ridiculous. This can easily house a number of continents larger than indie as well as an archipelago of *hundreds* of affliction sized islands. The ridiculous thing is the animal count of 50k. If we say half the map is water then there will be roughly 24.4 animals per kilometer of land. One square kilometer is 1 million square meters. So yeah... if 1 tile is 4 meters by 4 meters or 16 meters squared that means about 1 animal per 2,560 tiles and that's under the assumption that half the map is water. If like the rest of the servers it's mostly land then the wildlife will be even less dense. Good luck finding a pair of breeding animals. Going to take you weeks or maybe months of searching.

But since the wildlife is so spare this is effectively the peaceful server. Odds are if you don't move around much you'll be able to go weeks or maybe months without seeing an animal that isn't domestic.

I get the feeling when this was being planned someone didn't do the math. To get anything near reasonable animal density the number of animals that need to be on the server will need to be increased by several orders of magnitude and the way servers already buckle under the animal load I just don't see that being possible without disabling animal AI entirely and effectively turning them into animated objects.

 

Just my opinion, I am not Staff nor pretend to be. From what I have read on forums and news This is what I believe to be the case so far:

 

Why the stress test today? This isn't the final product, the server we will see today called Xanadu. This is a server in development and the stress test today is one step in a long process to get a big server like this up to the number of players and animals/mobs we are used to on older servers. Since Players are one of the most important elements that must be able to exist on a Wurm server it only makes sense that the Devs have to make certain at an early stage of development that Players can in fact log in and function on this kind of server or all the rest of the development is a waste of time. All they need apparently is one hour to determine this.

 

Xanadu isn't finished yet. Today's stress test is just to see if the underwear will stay on without bursting before we put on the corset and the dress.

 

This is what I assume is going on and I can also guess in the future we may be called to do another stress test with creature counts much higher and more "normal" like the other servers.

 

I'll be at stress test today to do my part to get this step of the process completed.

Edited April 10, 2014 by Kyrmius