#folia-help
1 messages · Page 39 of 1
paper-world-defaults.yml by @azure cloud: https://pastes.dev/Gsqm654MpS
paper-global.yml by @azure cloud: https://pastes.dev/iywxZlwYIU
check that the legacy enderpearl setting is set to false
but other than that, not everything works on Folia, thats just kinda how it is
folia works differently than paper
you cant expect the same things that work on paper to work on folia
I'm surprised that works in vanilla to be fair, the ender pearl should load 3x3 chunks, not just one
iirc it's forced to be legacy behavior on folia
Then is there a way to get folia to wirk like papper on this regard?
i mean, it was done for a reason
https://github.com/PaperMC/Folia/issues/318
you shouldnt ask here,
its not the place for asking fork supports.
ask the discord of fork.
very nice
and sinces canvas is a fork of that its probebly a folai thing
ofc but canvas have somethings about fixing things on folia.
if its not working, and also not working on folia, you should wait for a patch from leaf.
There isn't going to be a patch - as the issue mentions above that Tony linked, this is working as intended.
i was thinking about pearl stations...
We don't support forks here, period. But yes, if you reproduce the issue on Folia we can look into it often. However, as said above, this is working as intended.
If you're using a fork then in the future you should be seeking help from that fork. Folia forks especially tend to change a lot of stuff that we do not consider to be particularly "safe", so we definitely can't support forks here.
same logic on here?
This is the support channel for folia?
for folia itself, not for specific forks.
if u have issues on a fork, try upstream. Upstream fails, come here and say the upstream have issue
Yes, for Folia. We don't support forks, though.
yeah and folia is the problem
Sorry, I think we're getting our wires crossed here.
There are two different things. First, the thing you reported, which as you can see in this message #folia-help message is working as intended.
Second, is me reminding you that since you're running a fork, you'll want to seek help for other issues in the future from the fork you're using.
he reported it to me(Canvas dev), i told him to ask here if he could replicate on Folia
so he did come to me first, i redirected him here if he could replicate on Folia
i figured it may not work, but to be safe I wanted to check Folia incase i did cause it
(bc if i did i would obviously wanna fix it)
hopefully that helps clear stuff up
There's nothing to clear up. Once again, we understood what they were asking initially.
Once again, we don't support forks here. If they're running a fork they should seek help there. If you're maintaining a fork please don't send folks here for support. If there are issues with something and you believe it's us, have them reproduce it on Folia/Paper.
You can't control anyone coming here and not doing that. Which is why we reminded them that in the future any support they need shouldn't be here, because they aren't running our software.
We really don't want to spend any time on fork stuff in these help channels.
Which includes continuing to have this conversation.
i am trying to store some data into the pdc of furnaces in the FurnaceSmeltEvent.
the thing is when i call the update() on the blockstate that appears to cancel the event (no output item appears and no input item gets consumed).
when i dont call update() my custom data in the pdc gets lost.
is this a bug or intended behaviour and how can i work around this?
can you be a bit more specific, i.e show your code?
@EventHandler
public void onItemSmelted(FurnaceSmeltEvent event)
{
if (event.isCancelled())
return;
ItemStack result = event.getResult();
ItemStack source = event.getSource();
if (result.getType().isEdible())
{
if (source.getType().isEdible())
{
ItemMeta sourceMeta = source.getItemMeta();
long sourceTimestamp = sourceMeta.getPersistentDataContainer().get(FoodSpoilageTimeKey, PersistentDataType.LONG);
if (event.getBlock().getState() instanceof TileState tile)
{
tile.getPersistentDataContainer().set(FurnaceOutputSpoilageTimeKey, PersistentDataType.LONG, sourceTimestamp);
tile.update();
}
}
else
{
if (event.getBlock().getState() instanceof TileState tile)
{
tile.getPersistentDataContainer().set(FurnaceOutputSpoilageTimeKey, PersistentDataType.LONG, event.getBlock().getWorld().getGameTime() + (FOOD_SPOIL_TIME_DAYS * 24000));
tile.update();
}
}
}
}
this is my code and this causes the furnace to not smelt any items.
it will burn and the progress arrow moves but as soon as it is done the items dont get modified at all and it just "restarts" the smelting.
when i comment out the tile.update() then the furnace behaves normal again (but my data doesnt get stored)
what i am trying to do is to save a long value stored on the source itemstack onto the furnace blockstate
try removing the block state snapshots by using getState(false) and removing the tile.update(), using that inside that event is kind of known to have weird effects https://github.com/PaperMC/Paper/issues/4853
hey, does folia doesn't has any command like /data get?
Nope
they removed it i believe
should say it somewhere in their docs
/data merge would be a bit of a hassle to implement properly on folias structure and it basically doesnt serve any purpose in prod environments so its not worth implementing
once agian
i am trying to create a custom SmithingTransformRecipe and I noticed the constructor has the copyDataComponents boolean at the end.
as the "base" item in my recipe is a custom item that gets identified by a custom pdc value i set copyDataComponents to false.
now i have the problem that the result item (which i set to new ItemStack(Material.IRON_PICKAXE)) is still having my custom data and therefore keeps the texture (resource pack based) and name of my custom base item even tho i set copyDataComponents to false.
is that normal and am I missing something?
i hope this image helps you to understand my problem.
(the Iron Pickaxe Head is a minecraft:clock btw)
this is not a folia specific issue, i would ask about this in #paper-dev
Oh sorry... I see now, you have "Bot Spam" channel...
is there any way to avoid regions merging when players are close to eachother?
No
one of the invariants that folia relies on is that players close enough to "interact" must be in the same region
is that decided by view/simulation distance?
No, it's by the region size.
Which is somewhat dyamic.
If the regions are around 1000 blocks and a player gets too close they'll combine. That's just how Folia works.
If the players are too close together which'll cause regions merging into just one then Folia may not make sense for you.
makes sense ty
i guess you pushed to much into the folia repository that github pushs are now down
16 cores
you should need, if you want to get benefits. there is also a few things u should know about using it but first you should get enough resources
Maybe™
Now, we won't recommend that you try this as you're below the suggested minimum hardware
At what map size does folia make sense? If a server with 100 players and a border of 15k is it dooable or without a border and not pregenerate?
15k border size will probably not work for folia
I mean you always get the advantage of multiple worlds being ticked in parallel, but with 15k border and 100 evenly spread out players youll get a few regions max
Okay thanks
i think you can still get better performance than paper on that world.
it might not be the full potential of folia tho.
i am running a server with ~25 players at once (this is just a test, will later scale for more players) and all the players are within a 15k "border" (but i dont have the border set to that, they just didnt travel farther yet) and I still get ~6 regions (so 4 overworld regions + nether + end).
i still have 1 core at a higher usage (~80%) (depending on player count in that area ofcourse) so im obviously not getting the best use of folia but i think it would already overload that CPU core if i was running normal paper.
but in my case i have to mention that the CPUs of the server have terrible single core performance (2 Xeon gold CPUs).
So i guess Folia can make sense in your case if you have a bad single core performance but if your single core is good then paper is probably better in that case.
For context: my i5-11400F (6C, 12T) has more than twice the single core performance of the Xeon CPUs so on that CPU paper would make more sense
The difference between 25 and 100 players in terms of spread is enormous
What would be the problem using folia on a 16-thread?
Performance could be an issue depending on your server setup/player count/spread
What if everything is fine utilization, location of players, and number of regions?
Are there any other problems?
I mean, part of the recs to my understanding is a level of "this is how much you should be comfortable spending", using Folia without access to devs is somewhat of a dangerous game
You don't have to meet those specs to get performance benefits from Folia, you just need to be ready to deal with the caveats of a platform still in development in which mass support/adoption doesn't make too much sense
Hello, I am running a Folia server and I am attempting to increase the utilization value above the current 400%. I currently have 32 cores allocated to my virtual machine through proxmox. Is there a config option in Folia/paper somewhere to use more cores? I can attach logs/pictures if needed.
paper-global.yml
threaded-regions:
grid-exponent: 1
threads: 1 # increase this for more util.
@ancient jolt
I currently have the value there set to 15. Do I need to increase it further?
if you want?
How can I change the amount of regions that are created? I have a 1,500x1,500 play area, but it's still only making one region, total of two one for world and one for nether. I want to make like a new region every 30 chunks or something since the map is not an entire vanilla world.
tried all values for
threaded-regions:
grid-exponent: 1
threads: 10
threaded-regions:
grid-exponent: 8
threads: -1
only changes the utilization percentage
No, this map is tiny and you will always have only one region
That can't be reduced to allow more regions at all?
The bare minimum distance between two players for them to be in different regions is ~900 blocks
The map is too small for players to spread in such a way
And theres no way to change when a region is split?
Nah, you can't make it lower than that
Is there a reason? Is it unstable? Like modifying the Folia source code to make that a smaller region set is fine?
ShreddedPaper divides the region based on demand to avoid bad lag, you can also define these properties. So i'm assuming the best outcome would be like defining 8 regions or whatever, and it uses them for the worst area, rather than just statically defining the regions every 900 blocks, because if the whole server is within that one region like spawn, it's a bad time.
Is region transfering just not a thing? So that doesn't work? So it lazily results in just merging the regions.
its unstable because players can interact with eachother at that distance
folia makes sure that 2 players in different regions cannot interact with eachother (besidfes supported stuff like msg)
folia will strictly merge the regions whenever necessary (players interact) and only lazily split after longer amount of time
I can see that now. That method becomes extremely bad when all the players decide to bunch up.
yes you should just use paper if your setup does not fit the use case for folia
if you only have one region the regionizing is pure overhead
Wouldn't Folia still be a better option for adding multi threads to the regions? Or is one region literally a single thread?
you should read the writeups linked in the docs
it's actually all embedded now https://docs.papermc.io/folia/
folia is designed to scale when players are spread out, it's literally just not the right tool for what you're describing
Damn
My server had an influx of 3k people joining the discord within 2 days, our current playercap is 60 players. would it be worth switching over to folia. theres always around 100 people trying to join when its full
So long as you have a large enough world where people spread out enough and have enough cores to handle it then sure, Folia might be good for you
Plugins need to be explicitly ported to Folia though, it's not just a drop-in thing
Hello, i'm looking for plugin similary to RoseStacker for spawners and monsters (skyblock server)
Hello, please use google and search on site like Modrinth and Hanger for resources. This channel is for people who need help on Folia itself.
Don't found :/
how big is the overhead of Folia?
Like when i got a server that is the worst case for folia: a lot of players all in the same region, is there a performance difference in that case compared to regular paper?
The Nether and End dimensions doe get their own regions so at least those are still independant i guess and might reduce the load slightly but is that reduction in load bigger than the overhead?
I am asking because the server I currently run with Folia is getting very close to being single-core limited anyways and it might happen that a large group of players meets up.
The server has usually around 25 players online at once and I often only get 4 regions in total (2 overworld, 1 nether, 1 end).
The thread with the most usage is sitting at ~85% and I think merging the two overworld regions would cause it to be overloaded so I have to upgrade the system anyways (Currently using Xeon CPUs with lots of cores but bad single core performance).
So my question is: When I upgrade the CPU to have a much greater single core performance anyways, is it still worth sticking with Folia?
Using paper would be easier for coding my custom plugin and using already existing ones but I am planning the server to have +250 players at once in the future.
it is a civ-style server so most people will be in groups of ~20 people but people might still travel around visiting other civs.
Because of that I think Folia might not be the ideal solution, but even if I only get 2 overworld regions but both of them take more than 50% of a core that might still be better than paper, right?
The civs themselfes are usually at least 2k blocks apart from each other.
Think of my server as "SMP with a huge player count but 90% of players will always be in a bigger group and the groups are far away from each other"
Also my follow up question: If I decide to switch to paper, can I just replace the folia .jar with the paper one or do I need to reconfigure something?
Is my custom plugin which is 100% only tested on folia still compatible with paper without me having to change the code? (Do I need to port the plugin to paper?)
A single region in Folia is going to be somewhat slower than a Paper server but how much slower I don't think anyone has measured
Unless you're digging into some Folia-only internals there is I think one method that exists on Folia but not Paper, most likely your plugin would work fine on Paper
i dont think my plugin does anything special.
i think i will keep using folia if the performance penalty of the overhead is small enough.
like if it makes a performance difference of only 1 or 2 % of CPU on a few cores i dont mind that.
then i will still get the benefit of the nether and end being seperate and people traveling on expeditions far away being independant
Again, no one has measured, but my guess is that it'll be more than that.
If in general everyone is in one area and it's not a rare circumstance, it doesn't make sense to use Folia. Paper's going to have less missing bits and it'll be faster.
But no, you don't need to do anything. Just swap the jar for Paper.
Unless you made a whole bunch of config changes for Folia-specific stuff.
i did edit the thread count but i would guess paper would just ignore that
I have frequent timeouts and lag spikes on my server even tough I have good hardware. Can somebody tell me how I can figure out the cause of the problem? I would like my server to handle 200+ players.
@viral burrow you already been sent this on #paper-help not sure why you cross post here. We only support Paper and Folia here
Because this is a folia fork wrong channel
I have the same problem with folia 1.21.8 its not a folia issue
Do you know where i can get help then?
*folia fork issue
if you use folia, you can use this channel, if you don't use folia, you need to find out where you'll receive support for the software you're using
It would be the place where you obtain that server fork jar.
if u got a issue in a fork? try same issue on folia and once you got it, ask here for issue.
if you got a issue in a fork and didnt got any issues on folia itself? ask the maintainer of fork.
This was already answered, twice, like two hours before your reply. Definitely not necessary to ping someone to repeat the same message.
oh? then mb
Is there any folia supported skript plugin availabale ?
Likely no, it'd be against the point of Folia
If you find one that claims it does: it's lying to you. It doesn't. Maybe the server starts properly, but the moment any Skript code is being run, everything will break.
So i have to create my own plugin for that right ?
Really you just can't sanely make skript work with Folia
I'd argue whether you can sanely make it work period but with Folia especially it'd either have to hide a lot of thread switching and such from you or force you to write threaded code in the skript language
I don't know if that first one is even doable, probably not without a massive performance hit at least
You basically don’t want to use skript, even if you could wedge it in there for some stuff. Makes way more sense to use Java in plugins since you’re going to have to write a bunch of custom stuff anyway.
Is there a way/plugin to readd /restart for folia or add a bridge for Pterodactyl so UltimateAutoRestart can restart without manual input.
Restart is not a really complex thing to do
Sounds like something with that plugin is broken
Pterodactyl should also be able to restart without any issues
And afaik /restart was not removed on folia
You just have to configure it properly in the spigot.yml
the docs do contain it in https://docs.papermc.io/folia/faq/#what-commands-does-folia-disable
the /restart command is something not recommended anyways, on paper it's recommended to set the restart script to something that doesn't exist and let an auto restart script do it for you, on folia you'd just always use /stop
Why are those commands disabled?
They require a global state which really isn't a thing on Folia
Wdym "state"
He means since folia doesn’t have a main thread and ticks every region separately
Isnt there a global scheduler?
Yea but I don’t know much about it
the issue isn't that you don't have a global thread, the issue is synchronizing all existing region threads
Which runs on a seperate thread
Folia works by splitting up the global state of things like the world into reasonably isolated units of work
Things like the scoreboard rely on it being modified and read dozens of potentially thousands of times a tick, which can also then trigger off other modifications elsewhere
The only options there are to basically either disable the thing entirely (and let server owners decide what they want to do) or introduce a whole lot of complex locking and mechanisms for pushing out those updates into every single region for them to decide how they need to respond to it
Folia is really just meant as a base project to then have devs work on and tune to your liking
How to get folia 1.21.4 jar file?
See the build explorer on the website
Oldest one is "Update to 1.21,7"
Oh
I found the build explorer sry
You really shouldn’t run older versions. There are a lot of improvements and fixes in each version of Folia that aren’t related to Minecraft
yeah ik but im now making a plugin and i have to support folia and other versions
and im just testing it rn
But still i think folia 1.21.4 is better than paper in performance
- already my dev server is 1.21.4 cuz like i feel like 1.21.4 has the best features and its not that heavy unlike 1.21.8 it has a lot of new structures… so i like to use .4 always
i might try 1.21.8 but still i hate the new structures and i might make a new plugin that tries to remove them
If you're making a Paper plugin you're better off supporting 1.21.8+
alr but why exactly cuz a lot of ppl r like using paper versions under 1.21.8
Like I said, with Folia especially people want to be on the most recent version.
Most people are on the most recent version of Paper also.
you are right but u can see also there is a bit less percentage of ppl using folia 1.21.4
Thats why i will be supporting 1.21,4 and later
Just because people do something doesnt mean that it is a good thing to do
Hi, I’m on Folia and want to hide the server version/core in the server list. I have ProtocolLib. Any lightweight plugin or minimal solution to spoof only the server type/version? Thanks!
I doubt anyone will help you with that here
You don't even need protocollib for that
The brand is the only thing that "advertises" this software
my favorite thing on there is making f3 server brand "vanilla" lmao
I once had a Fabric server and I made it display Forge xD
what even is the use case for this
people like giving attackers as little information about server structure as possible
Fair but Folia is extremely easy to detect regardless lol
as mja said, it's extremely easy to find that out, for instance, most people know 2b2t uses folia lol
yeah but 2b2t doesn't mind people cheating / exploiting (or atleast not really)
yeah no doubt but id argue in an optimal case you want your server to be a black box
isnt everyone wants to make their server black box?
maybe not everyone, but its pretty common
i can not see open-source server setups on github these days, because minecraft servers is in gray area
And 99% of those servers fail to hide the information properly
and i think people dont want to make public their configs, everything
fr lol
perhaps
but im just saying theres most defenetly reasons why people may want to hide their server brand, even if the information loss from an attacker's side is minimal
I have a simple tool to extract every bit of information about a server that is available
That tool can very accurately detect proxies and server software no matter how hard you try to hide it
There is always something that isn't "protected" (hidden) properly
Knowing the server version makes you an easier target for serious attackers in my opinion than simply "paper" or "folia"
no doubt
i was doing same researchments on minehut, when im joining ppl servers instantly sending /bukkit:pl and /icanhasbukkit 😭
making it at least a little harder to find out might stop a few "script kiddies" tho.
Of course some "real" hacker wont be stopped by that but at least some kids that just copy-paste random stuff from github might see that as a roadblock
A good client will check them anyway regardless of what it says on F4
So it is mostly just cosmetic. It doesn’t offer any protection.
Is anyone willing to vc with me and help figure out some issues im having with load on my server?
I think I have a problem with folia, I would like to make an issue on it, but its a bit too abstract for me
We don't really do VC support, a spark report should suffice
What KP said. Also if you plan to open an issue about it, you're gonna have to write it all down anyway, so articulating what you want via text is a good plan.
Hi I’m new to the project. Can anyone point me to any resources / conversations about Folia vs MultiPaper vs ShreddedPaper? I want to understand what each is trying to optimize for in solving the single core bottleneck issue.
I’m a distributed systems SWE working at a large company who wants to help contribute to solving scaling issues. My cursory understanding is that proper horizontal scaling is theoretically the best solution for this problem. But I understand that Minecraft has many constraints that make horizontal scaling hard to get right (I.e duping due to consistency issues).
Folia is the only implementation that seems to stand so far
ShreddedPaper is in beta, MultiPaper is known to be pretty buggy
Thanks KP. As I understand it, Folia solves the issue by being vertically scalable?
Yep, the scaling is "proportional" to the spread of players
The core mechanism being isolating different regions' states which can then be ticked in different threads
but when two far players going near, their region merged
how do i prevent seed cracking (for xray) on folia?
packet
@uncut pelican ^
Someone have a link for FAWE forked for Folia ? With compatibility to 1.21.8 ?
thanks
i ve to clone and recompile ?
yes, i can send you the jar in DMs, but I can't do so here, because PaperMC team does not allow sending random jars, preventing viruses etc
it's always best to compile it yourself, there's usually a very easy tutorial
FWIW we don’t want people sending jars in DMs, either. That’s kinda worse. We don’t want to normalize people grabbing potentially malicious jars from people they don’t know.
oh yeah, thats completely understandable, so we should just tell people to compile it themselves if it's opensource?
and what about when it's not
Yeah exactly
When it’s not open source they should get it from the official place that distributes the jar. Hangar, github, whatever.
what about if it's a private fork of a plugin that added folia support, for instance
If it’s a private plugin, it should have a place to download the jar.
If it doesn’t, then you probably shouldn’t be distributing it anyway.
doesnt ore generating have a separate seed ?
what if they find that seed?
that's how it works yes, and if they find the seed despite all this? kudo to them, they will break your custom generation anyway if they can figure out subseed.
how likely is it that they will find ore seed?
average MC player? zero
any mods that already do this?
will they need big chunks of world data to do this?
You can write up a custom script and bruteforce it, do you think your player will go that far?
lol it will probably take a month
I don't know how much data you need from ores alone to find the seed, but I suspect it's a lot.
That's why all the seed cracking uses structures.
alr, thanks. i will use the feature seed thingy and change structure seeds in spigot.yml
similar feat is this https://www.youtube.com/watch?v=GaRurhiK-Lk and it took them months but they are able to get it via a rough screenshot, the point is this is a balancing act of effort vs reward.
The Minecraft title screen. We've all seen it. But have you ever wonderered if a seed existed for it?
Minecraft Beta 1.7.3 Seeds:
2151901553968352745
8091867987493326313
Image Overlay Program: https://codepen.io/Tomlacko/pen/mdVOOwG
Contributors List: https://docs.google.com/spreadsheets/d/1RkNSNfmrDdDH_yYTKE7MVA1mgjuqN9HYWE7vYJh-UPQ
Earthcomp...
Be aware that you can only do this on a new world.
Changing this won't do anything on an existing world.
You have to set them and wipe the world / generate a new one.
btw what options do i have when upgrading my world to newer versions?
where are these seeds stored?
What do you mean? Paper/Folia will automatically upgrade a world.
But for seed stuff? Like I said, you have to delete the world and start again.
lets say i generated the world with this seed feature on 1.21. then 1.22 came out
can i easily upgrade my world?
Yeah, that's not a problem.
👍 thanks
any eta on stable folia 1.21.10? im failing to get worldguard (folia version) working on 1.21.8
We don't have ETAs for stuff.
(So no)
is the current build stable enough? (if i build it myself)
(assuming there is already a version for 1.21.10 on github)
There isn't a 1.21.10 branch yet. But if there were a branch and there weren't builds then it wouldn't be stable enough.
U can technically edit the world dat file
But u gonna get chunkborder for new chunk
It ugly but possible if u wana go that route
can i run a folia server giving all cores i have + other servers, even tho folia may not have many players at start but hoping it will one day and because players will be making crazy farms and setups paper server will start to lag after 30-40 players?
7900x 192GB Ram
Wdym + other servers?
What are you really asking?
1 survival server
total 2 servers, 1 with paper where i will have 40-50 players on it
1 other server with folia 10-15 plugins and semi vanilla setup maybe 50-100 players
can they both share the same resources as both on same machine
Yes?
If you got enough cores you can "reserve" some for the other paper server sure
But you dont have enough cores
A single folia server alone wants at least 16 cores
I mean it still works might just result in worse performance
oh i only have 12 cores, 24 threads
It might not be reccommended but you can still try
Folia will probably be faster than paper on your machine, but by how much and if its worth it with plugin compatibilty is another question
Yea, you wont get the full potential of folia tho
But might still be good enough for you if you got very good single core
On your machine it would probably still run better than on my 24 core xeon server
7900x should be decent! 4.7ghz and boost is 5.4ghz
It is not recommended but you probably will get better performance than with paper
its been 8 years my server haven't grow more than 70-80 players because of lag on my server. So trying out other game modes and setups hopefully it grows
the suggestion is minimum 16 cores, not threads but you can use it if you know what ur doing.
It will perfectly work even with 2 allocated threads
It wont be "perfectly".
That is simply not how it was designed
Explain
What was the design so it won't work with 2 worker threads
You will not get the full performance.
At a minimum folia has 1 region per dimension at all times so there are at least 3 threads used at all times.
Overworld, nether, end
So yes it will work, you will just run into single core bottlenecks sooner or later
I mean, it will work in such situations, you're just going to inherit a lot of the negatives of folia for potentially little gain, and potentially much hurt, depending on the environment, etc
so no i guess for production server unless you have 16 cores or higher
paper server i could never cross more than 50 players without dropping tps to 15-16
vanilla/spigot, others are impossible anyway
You can do it with less
Like, for me, the expectations over hardware are more a "resource investment bar" than anything else; You shouldn't be using folia unless you have access to devs able to work on stuff, etc
is it normal that this one region has 13.000 chunks? (render distance is capped at 10)
Hi, I have an issue with a player. The server says that he's already connected, but he isn't. There's no ghost of him. I can teleport to him, but even kicking or banning him doesn't work - the server still says he's online when he actually isn't.
With that many players in the region yes, that can happen
Any chance to force the server to split those players
I'm currently running on 1.21.4 version
because they are far apart but still in the same region
You can reduce the grid exponent but that will only get you so far.
If the players form a chain so every player only has 2 other players within the tegion range then they will all still be in 1 large region
What grid-exponent would you recommend? Never had that problem so never changed it
threaded-regions:
grid-exponent: 4
threads: -1
My current default config never changed it
Honestly dont know rn, but there probably is some better explanation for it online
OK but thank you very much
You can only make regions so small. At some point if players get close together they’re going to combine. That’s just how Folia works. IIRC the minimum is like 900 blocks.
lowered grid-exponent by 1 and told players to move away more. Server just started hope it helps
i dont think you will get performance from shenandoah. Its generally not designed for minecraft. Just take a look on ZGC.
If i remember correctly, one thread will often be responsible for ticking multiple regions
well, crashes how? the flag shouldn't really be a factor for folia, if the JVM is crashing, you'd want to look down that road
i use it but it dont works
openjdk version "21.0.4" 2024-07-16 LTS
OpenJDK Runtime Environment Temurin-21.0.4+7 (build 21.0.4+7-LTS)
OpenJDK 64-Bit Server VM Temurin-21.0.4+7 (build 21.0.4+7-LTS, mixed mode, sharing)
STARTUP /home/container: java -Xms95G -Xmx95G -XX:+AlwaysPreTouch -XX:+UnlockDiagnosticVMOptions -XX:+UnlockExperimentalVMOptions -XX:+HeapDumpOnOutOfMemoryError -XX:+UseLargePages -XX:LargePageSizeInBytes=2M -XX:+UseShenandoahGC -XX:ShenandoahGCMode=generational -XX:-ShenandoahPacing -XX:+ParallelRefProcEnabled -XX:ShenandoahGCHeuristics=adaptive -XX:ShenandoahInitFreeThreshold=55 -XX:ShenandoahGarbageThreshold=30 -XX:ShenandoahMinFreeThreshold=20 -XX:ShenandoahAllocSpikeFactor=10 -XX:ParallelGCThreads=10 -XX:ConcGCThreads=4 -Xlog:gc*:logs/gc.log:time,uptime:filecount=15,filesize=1M -Dchunky.maxWorkingCount=600 -jar server.jar
[0.002s][warning][pagesize] UseLargePages disabled, no large pages configured and available on the system.
Error occurred during initialization of VM
Unknown -XX:ShenandoahGCMode option
pelican@Folia 1.21.4 ~ Server marked as offline...
I mean, is that not just a standard Temurin JDK that they're using?
his said that hes using shipilevdev's java 21 shenandoah
and I'm a cat, nyan
Nightly builds of OpenJDK 8u, 11u, 17u, and other related projects. Fastdebug and slowdebug builds for improved JVM debugging. JDK backports monitors. Java Mission Control binaries.
and i mean, if ur using dedicated alr, you shouldnt use webpanel
i hate webpanels if i would going for more
So, yea, double checking, they're 100% not using the shipilev builds
openjdk version "21.0.5-testing" 2024-10-15
OpenJDK Runtime Environment (build 21.0.5-testing-builds.shipilev.net-openjdk-jdk21-shenandoah-b8-20240820-0053)
OpenJDK 64-Bit Server VM (build 21.0.5-testing-builds.shipilev.net-openjdk-jdk21-shenandoah-b8-20240820-0053, mixed mode, sharing)
Then no idea unfortunately
Is there an alpha or test build i can try for 1.21.9 ?
Nope, just 1.21.8 for now.
There won’t be a 1.21.9, anyway. Either 1.21.10 or right to 1.21.11 possibly.
Alright, thanks
Hello. I often encounter these errors. Can you tell me what to do about them?
Ye, it does seem to be a core thing
You get it on random coords or always around the one you are showing?
The coordinates are different in different errors.
Ye, expected that, but was hopping for it to be region based
how can i use folia with prism launcher
Prism Launcher is for Minecraft clients as far as I know. Folia is Minecraft server software, you do not use Apple for Orange.
alr
What's the permission node i need to give someone so they get access to the /tps command?
yo can someone help me, with folia i have the error: POI data mismatch: never registered at BlockPos
when i get the block region thread in a blockplaceevent..
i need to run this during server shutdown
but the api doesnt let me do that
how can i do it correctly ?
You cannot use schedulers when the plugin stops
so how can i achieve that during server stop ?
I don't know if it's "possible," but when you do /stop, listen for ServerCommandEvent, cancel it, close the inventory, and stop your server on the next tick (you can also kick the players)
Hi, I think I have a memory leak on my folia:
[14:21:00 INFO]: Chunks in world:
[14:21:00 INFO]: Total: 872104 Unloadable: 21918 Null: 567628 ReadOnly: 171 Proto: 0 Full: 304305
I've done a dump and I see many instances of the ca.spottedleaf.moonrise class.
872k chunks, interesting xD
xD
you have a pretty nasty leak, but it's not because of Folia, but shitty plugins
What worries me is the amount of Null there is.
I don't have a lot of public plugins, and in the dump I can't find any references other than paper, spottedleaf, or minecraft.
which version of folia are you using
we've been using the latest 1.21.8 without any issues
1.21.8
I can't paste the logs here, there is a lot of information from many users
the site provided has an option to strip IPs
cant really help you without any info provided
It tells me that command does not exist.
You would probably want to dump the chunkholders info
[14:48:35 INFO]: No force loaded chunks were found in minecraft:overworld
[14:48:43 INFO]: No force loaded chunks were found in minecraft:the_nether
[14:49:01 INFO]: No force loaded chunks were found in minecraft:the_end
We've been trying to fix this error for weeks and haven't been able to solve it
when obnDisable is called, all players will have already been kicked from the server
pretty sure atleast
No, but because it's really just getting started. After a few hours, the memory fills up and the GC goes crazy and freezes the server at times.
can you make a heap dump
actually that wont be all that usefull
Sure! I've already done it. Let me open it.
Oh, okay, I'll do it in about five hours.
but I probably won't be able to open it 😂
If that's what I'm doing to open this 64 GB dump 😂 😂 😂
yes, but it has little uptime and no players (because I kicked them out)
Do you think it should be with the players inside?
but I think this command does not exists
Yes, but if it's a memory leak, I suspect that with or without players, it will still be there. I wouldn't want to leave the server frozen at 1,500 players 😂
like 5m writing in the disk hahaha
😂
well, you have to
isnt it possible to write some code that just dumps whatever tickets are keeping those chunks loaded?
have version in 1.21.10?
The problem is that I used version 1.21.8 Folia and couldn't get WorldEdit or WorldGuard to work with it, but I managed to get them working for version 1.21.10 Folia.
folia 1.21.10 does not exist
Okay, do you know if there's a working version of the WorldEdit and WorldGuard plugins?
no, I don't think they officially support folia yet
Oh, that complicates things, I don't know what to do. So, are there any plugins that do the same thing for Folia?
Last I knew there are community forks of those types of plugins
ofc, you're not going to get much support for them, however
Yes, I searched but found nothing, and they didn't work on version 1.21.8; it was asking for 1.21.10, which I found strange since that version isn't available.
i know it’s crazy
i mean you doesnt even that much allocation
pure waste
for even 1000-2000 players*
my server has 96GB.
The peak usage (of the whole server reported by the OS) was 13GB xD.
I just wanted 12 RAM modules to get the 12 memory channels, didnt really care about the capacity that much.
i guess(lol) its all about how people made farms, things when the time pasts
to be fair my server only had 33 players at the peak.
But I do have a high simulation and render distance and my custom plugin is not very efficient as well xD
guess my statement is kinda useless as the numbers dont compare lol
Hi, my server on folia have around 200 players, and have multiple massive pigman mob farm running with no mob cap limit. My server CPU is 9950x, I would like to upgrade the cpu to something else.
I checked this cpu benchmark list https://www.cpubenchmark.net/cpu_list.php
And I saw 9980x with 64 core 3.2-5.2Ghz have a cpu mark similar to 7995wx with 96 core 2.5-5.1Ghz, but 9980x is much cheaper, may I know if their performance will be similar if I run folia?
Or 32 extra core from 7995wx really matters a lot
it would be beneficial if you could share a spark report
that way, we will know what exactly needs upgrading
for instance, if you have memory problems, or plugin problems, upgrading a CPU will not help you
Thanks, I just took a spark report and it show an error occured whilst uploading the result. I can only display it locally 🥴
400MB can't upload here too
We don't provide support for software that isn't ours
there are many things you can optimise regrading paper/bukkit/spigot configurations
feel free to dm me if you want me to help you out
but i don't want to optimize these things as i want to keep vanilla behaviors, thats why i am asking for better hardware advice 🥴
there is a lot of question marks on there and luminol is not a product of papermc, its a fork unfortunately
you can keep vanilla but there is a cost lol
Please keep this channel on topic, like Cat said, if you are not using our software, please seek help from the maintainer of the fork you are using.
@narrow tangle I ask the topic to be stopped.
I understand and I am asking for folia, why everyone is judging me for other forks?
and no fork is not still Folia, that's by defination not true.
because you are not using Folia
you are using a fork of Folia.
We can only provide support for software that we make.
I understand, i am just giving an example of spark profiler, I have been using folia without a fork for almost 2year at least since folia was born
I am a folia supporter since the start of this project in asia region
Then show us your problem while you are using Folia and not a fork if you want to use our free community support.
This is a very simple ask.
ok
yo
yo does anyone know how to make it so [member] is the default rank?.
so anyone who joins insta gets that rank
luck perms
i guess this is not a good idea for asking this channel.
there is something i want to ask
let’s say you are player a and you want to send money to player b
if you move from a’s region to b’s region and perform the money transfer there the operation will run asynchronously because there is no lock between the regions
this can cause the same transaction to be executed at the same time by different threads leading to inconsistent results is there any built in protection for this in folia or do we have to handle it ourselves
if you have 2 players interact with each other you should always use the player's schedulers.
doesnt matter if they stand right next to each other or are 100k blocks away that way
that is, it should not be async?
two processes = one runtime
if you call a different scheduler to do something that will probably happen on the next tick then.
so it is async
whenever you execute scheduler.run or scheduler.execute or whatever it will always schedule the code to run on the next tick if i remember correctly
if the next tick uses the same variable then the data will not be deleted before your code is done with it
easiest solution would probably to handle that outside of tick threads completely. i.e offload the transaction onto the global thread
like microservices?
with redis lock
if were thinking of microservices in the same way, then no, thats not what i mean.
I mean if you have something like money per player, youd want to store that data in a central database which you only access from one thread. so you dont do
thread 1: take player A's money
thread 2: increase player B's money
but rather run the entire logic on the global scheduler (or maybe even an async scheduler) where you can do some locking
but I think youd need a very weird setup to get that problem in the first place, unless youre storing your money directly in player data
thanks
Global scheduler is more logical than that
I mean, folia has 0 care about your plugins logic, any synchronization/thread safety you need to deal with yourself
Just throwing all of that stuff onto a single thread is often the easy way, especially if you don’t have any contention issues, otherwise, boils down to design, etc; if using a database, they generally have some good tools for making this easier
will there be a 1.21.10 folia version?
more logical than what? what i explained there would work on the global scheduler
More logical than using ("random") one of the players' schedulers
but i explicitly mentioned not to use one of the players schedulers, i suggested using the global one
I assumed a broader context reference in the answer from yunuservices but yeah, looks like it lead to more confusion than I hoped to solve.
is there anyone here free and is smart at authme?
i need to make it so when you join you get tped to spawn but when you log in sucsesfully you get tped to ur last place*
dm me
are you running an offline-mode server?
yep
thats why i have auth me...
i needed help on making auth me auto tps to spawn
you wont get that help here, good luck
not a note saying papermc doesn't support cracked
offline-mode = cracked
u sure?
no sh* sh*rlock
so u are flexing that you break the rules rn?
good for you if you got the help you need.
i dont see why you have to show a screenshot of that tho
papermc are breaking the rules too by supporting it
you said i won't find help here so i showed you that i can
ever heard the term "debug/dev environment"?
that person is not on this server so you did not find that help from this server
Mojang also has this feature 🫢
people dont have a account
Comeon 27$ one time purchase
if they really like the game they should buy it (Unless they really can't afford it )🤷♂️
I also had an offline supported server, only made our security worse.
We were worried about the player count but we also noticed its really illegal and can get your server shut down by Mojang...
Yeah but the player count rose afterwards 
@coarse zephyr I've mtued you for a day. We don't support offline mode here. Please see our rules in #welcome - specifically rule #7.
Everyone else - just say the thing once and move on. You don't need to argue with this person.
[ca.spottedleaf.moonrise.common.util.TickThread] Thread failed main thread check: Cannot asynchronously load chunks, context=[thread=Region Scheduler Thread #5,class=io.papermc.paper.threadedregions.TickRegionScheduler$TickThreadRunner,region={center=[917, 864],world=world_italia}], world=world_italia, chunk_pos=[0, 0]
java.lang.Throwable: null
what's that T-T
that is an error
please provide more information about your server, otherwise we cant help you
this error just happens and the server crashes, it's been going on for days
the most important questions for starters:
What version is the server?
exactly when does the error happen?
does it also happen without any plugins?
1.21.8
idk, random
idk, idk how to replicate it
and some general context about your server would be hepfull:
is it just a small private smp or a bigger server?
I have a server with about 200 players
have you tested it without plugins and/or datapacks?
does the error only happen when players are online?
I can't try without plugins because I don't know how to replicate it, it happens randomly so I wouldn't know how to replicate it (and I can't open the server without plugins lol)
run the server for a few hours without plugins and see if the error happens
yeah, i always have some player online
is there a crash log file you can share?
i can send the entire error
please do
if you send a file, make sure it doesnt contain personal information that you are not comfortable with sharing
Pastebin.com is the number one paste tool since 2002. Pastebin is a website where you can store text online for a set period of time.
it's like this
https://github.com/PaperMC/Folia/issues/392
Chestprotect plugin is most likely to causes the error
mh
try removing that plugin and see if the error still happens
the technical explanation (that you would want to report to the developer of that plugin if that ends up being the problem) is that it does something that is not thread-save on the wrong thread
seems like there is some code obfuscation going on so the error can not give more detailed information in which function the error happens
mhm, i'll try
I will contact the developer
please make sure that the plugin is the actual problem before you report it
the error also lists some other plugins so it is possible that it is not the chestprotect plugin's fault (even tho it is most likely)
for now just remove the plugin and if the error doesnt happen anymore for a few days you can report the problem
Also, you're using 1.21.4, not 1.21.8
We don't really support 1.21.4 anymore. Would recommend at least updating to 1.21.8
Oh. And you're in offline mode. We don't support offline mode - which you've been told.
where did you read it?
this is in an online mode server
Send a spark report of the server?
it happens with 1.21.8 too, it doesn't improve the situation bruh
Like I said - send the spark report.
I don't find it necessary now, I'm not asking for support at this moment
You're claiming that your server is not in offline mode, but you continue to ask for support for offline mode servers. If you ask for support here again you may be banned. Please see our rules in #welcome - specifically rule #7. We don't support offline mode servers and don't really want to waste time with people claiming otherwise.
bro before I never said that the server is in offline mode, I did it in the past and the situation has changed
but ok
bruh
How does folia handle regions that are hung?
What do you mean with hung?
i never trust ppl starting an argument with "Bro"
Make it just kill them and delete the affected world file. As an added bonus it prevents the issue in the future by reducing load by causing the players to quit after losing all their progress.
it doesn't handle them
If you actually have 1,000 players available to join, which is unlikely, you're going to need a CPU/server that costs an absurd amount of money. Search this channels for previous tests and system specs for large players.
Also be very aware of this:
You aren't going to fit 1000 players in one small area.
But do u have a CPU in mind?
Is AMD Epyc 7642 a good CPU for folia 1k players?
Folia needs player spreading and at least good specs according ur requests.
You should read old messages in: #folia-help and #folia-dev
I do not have a CPU to reccommend to you. As I said, getting 1,000 players on one server is going to be very tricky and require heavy hardware. I'm not sure someone could just slap Folia on a box and turn it on and get 1,000 people. You'd have to configure things for high player loads and have people spread out 1500+ blocks apart like the image says above.
If you don't already have a 1,000 player server, start small and grow the playerbase. You'll learn how to run a larger server load in time.
1000 players, i recommend amd threadripper 9970x
hello everyone, does orbital strike works in folia?
not the orbital strike cannon 😭 😭
try and see i guess
im just too lazy to try this in survival
mb some settings disable it
okay cool i guess
theres just a problem
i want to play mc with my friend, but he doesnt want to
BUT he loves orbital strike cannon and he wants to build it sm
start a local server past an orbital cannon with a schematic and try it
easy
no its just looks cool
naw
were playing on a smp
he meant do that in order to figure out wheter it works
lol
GIFs and memes go in #gifs-and-memes
Also if you’re just playing with one friend you wouldn’t use Folia
oh ok
You’d want to just use regular Paper
no, were playing on smp, ive said that before
Ah gotcha. I missed that part.
This will not work on Folia.
thats sad
What’s so sad about it? Can you explain?
thanks for the answer!
.
You should be using Fabric server software if that’s the case
You do not need to run Folia for a server with friends playing together. It will not be the most ideal.
message.txt by @dapper field: https://pastes.dev/wZzodV0P8t
how to fix its
Im running a Folia server on Windows Server. Is there any way to force players to join using my domain name only and completely block joining via the IP, without using a proxy likee Velocity or Bungee?
there are plugins for that
i use tcpshield but there are probably other plugins that dont require any setup like that
just look through hangar or modrinth
This server is running Folia version 1.21.8-6-ver/1.21.8@612d9bd (2025-09-30T14:11:15Z) (Implementing API version 1.21.8-R0.1-SNAPSHOT)
You are running the latest version
Previous version: 1.21.8-DEV-1958fdc (MC: 1.21.8)
[15:25:33 INFO]: Paper Plugins (6):
[15:25:33 INFO]: - CoinsEngine, FancyHolograms, FancyNpcs, nightcore, SmartSpawner, Worlds
[15:25:33 INFO]: Bukkit Plugins (61):
[15:25:33 INFO]: - ajLeaderboards, AnarchyExploitFixes, AntiCrasher, AxInventoryRestore, AxSellwands, AxVaults, BetterRTP, Chunky, ClickableHeads
[15:25:33 INFO]: CustomEnderChest, CustomJoinMessages, DeluxeCoinflip, DeluxeMenus, DonutCrateSystem, DonutDatabase, DonutKeyall, DonutOrder, DonutSettings, EnhancedStats
[15:25:33 INFO]: Essentials, ExploitFixer, GiftCode24, GrimAC, HamsterAPI
[15:25:33 INFO]: justTeams, KnockbackSync, LagFixer, LeaderboardMaker, LiteBans, LPC, LPX, LuckPerms, Matrix, McRtpQueue
[15:25:33 INFO]: MythicTools, NBTAPI, packetevents, Pl-Hide-Pro, PlaceholderAPI, PlayerKits2, ProtocolLib, PvPManager, SkBee, SkinsRestorer
[15:25:33 INFO]: Skript, spark, TAB, Terra, ToastedAFK, TotemGuard, Vault, ViaBackwards, ViaVersion, WorldEdit
[15:25:33 INFO]: WorldGuard
The chunk system stalled out
you would at the very least need to provide a jstack and maybe full logs
Cracked servers are not supported
not cracked
why use SkinsRestorer ?
message.txt by @dapper field: https://pastes.dev/3dEc7oNvXK
I mean, so the thing is recovering, it's just getting stuck for some reason
block error causes the entire server to crash, killing everyone and the server cannot be stopped unless killed
You would need to work out why it's locking up; outside of that, that network is in offline mode, so I'm generally out
what plugin do u suggest if yk any
i dont know any
thats why i said to look on hangar or modrinth
is he cooked?
I think i found (probably a known) nasty bug with entities going through portals losing their source. (wither shooting a skull through a portal doesnt drop wither roses when an entity is killed on the other side)
Hey! So I currently have 4–7 nodes with Ryzen 7950X CPUs running cross-server Minecraft. I’m planning to switch to a single 9970X machine with 256GB of RAM to handle 1,000–2,000 players. Would this work well?
uh.
That's going to be interesting, can't really answer how well it'll do at that scale but you'll have to make sure you keep the player spread, be it across worlds or one world
one world, sharding?
to single instance
i mean, you may handle it but it would cost a lot of money atp
yes, basically split the map and assign each part to a separate instance.
yes i know this
not a good pratice of
sharding but yeah for now best way to do it
Welp, my server is really huge, like 7tib of world data
💀
xeon what
uh just cpu name
bros flexing to us with his homelab?
damn
anyways jokes aside
u may handle it with a single instance
these amount of player counts
ah ok i hope folia stable now 🙏
well as now im using folia it using like 20 thread of 7950x to handle 300-400 players
is folia out for .11 pre release?
Hello!
I'm using CombatLogX ( plugin supported by folia ) and the timer and the text is on the scoreboard, why does my sccoreboard not show up after hitting them? ( it works on paper, not on folia tho )
i am new to paper and folia and all this, im trying to learn everything
How could I implement it
How would that work?
ProtocolLib does it for u already,
?
I solved it with the tab plugin scoreboard feature
how?
Read the wiki
In folia, the scoreboard is not implemented yet, so you have to use packets.
I deleted all my plugins and the villager still doesn't recognize the bread and therefore it doesn't reproduce. I'm using version 1.21.8.
Maybe it's inventory is full
I don't think so, since I made two new villagers appear and they still don't get the bread.
Do you check your gamerule
Does anyone know how to help me? I've already tried it without plugins and it still doesn't work.
Please perform your test again with no plugin/datapacks and fresh new Villagers. It is likely you missed one part of those factors in the test (as seem in your screenshot provided to us, you still have plugin installed).
Yes, I put it back on, but it doesn't change anything. I'll send another video.
Remember that I did this with all the folders deleted and let them regenerate normally. I deleted everything, everything, and it still doesn't work.
don't get me wrong but how can we be so certain?
The server was completely wiped clean, I deleted all the folders. GameRule is correct, I already checked and was told about this before.
Do you have the same problem?
@mortal burrow We don't allow paid services / soliciation here.
Please tell me you calculated too high 😭 🤞
Seems the best solution is increase border, pre render, and set the random spawn radius to whatever the border is.
A visualization of a 50k worldborder map (r=25k) with 120 players randomly placed in it
The hard limit for two players to be in different regions is around ~1k blocks, this is made with a region of radius 2.5k to give them a ~1.5k leeway to walk around without merging. This situation creates one big region
I had messed up radius vs diameter and ended up doubling what I meant to say
So every box that’s touching, is that connecting a region?
Now this doesn’t account for players living in the nether or end correct?
those will have their own regions, yes
Cause when I previously used folia my big issue was “78 players in X region”
So the better you can do of distributing players the more you can handle is the verdict?
Yeah, they would be one region (if players moved around enough for them to connect)
Is this a website
I’m curious how 100k looks
Yes, if your hardware can handle it (single threaded performance for ticking heavy regions, multi-threaded perf + core count for ticking the regions)
This is vibe coded and I'm just verifying if the outputs are correct before sending, lol, I can DM it to you but trying to not publish it for now
Yeah I have no intentions of sharing it around
100k looks a lot better
Cool cool
I'm pretty sure even the non-touching ones would be merged in this scenario
just because of how the radius aware locking works, anything in the square about [minChunkX, minChunkZ] -> [maxChunkX, maxChunkZ] will be merged
but I might be remembering wrong
Well, the squares are meant to represent a generic simplified "movement buffers" rather than what the regionizer works with (as in, "you can move somewhere around here until you eventually cause some merging")
is it safe to switch from folia to paper?
Yep.
yeah I was wrong anyway, was confusing the regionizing with the radius aware locks which aren't acquired for an entire region
I use folia and I do not have the issue
Even with a billion plugins
What is that website im so lost 😭
ur funny.
Please don't post joke/troll answers in help channels.
Sorry
wdym paper builds for 1.21.11 are already out but we're still waiting for folia 1.21.9
Unlike Paper which has a team of developers behind it, Folia is largely a solo project with a bus factor of one, there will be no ETA on release, you can subscribe to the #announcements if you would like to get notified when it is released @brittle crag
alright appreciate that
Aprox. its gonna be out something like 2-3 months most likely
you don't know that
Ofc he doesnt know but previously it was 2 months wait for 21.8
Let’s not making up numbers or timelines
More like disappointed 
Angy
makes sense
I’ve figured out that folia 1.21.11 will release in 1T
-# T = the time required for folia 1.21.11 to release
Do you think 96G will enough for Folia?
with 100 plugins to handle like 300-500 players on 6.7tib map
why is this only 400
You dont give any info is it 96G of storage ? or ram ? if its ram is it DDR3 DDR4 or DDR5 if DDR4 and above then ram shouldn't be the issue
paper-global.yml > threaded-regions
Section of threads:
Does anyone have the same problem with folia 1.21.8 latest that the ping is like very high if i switch to paper its fixed
thanks, do you recommend what I should increase it to? I have 32 threads and I'm assuming putting all 32 here is a bad idea.
putting all of it to here is ofc bad idea
You should enter values based on server supply and demand.
96G DDR5 RAM
didint i already said 6.7tib map (mean storage) lol
also any know to fix this?
I'm using chunky to pre-gre but only 200-240 cps
specs: 9950x, 192g
Thread Allocations:
GC: 3 concurrent
Chunk System IO: 2
Chunk System Worker: 1
Netty: 10
Region Threads: 12
Should handel about 500 players at 20tps when spread over 50 regions
If u spread the players well over 30 regions it should work. But for what do u need 100 plugins?
because is an earth server lol
not like smp
180 threads cpu for 2000 players is nice (overall network)
Ohh ok ok, ya when u spread players it should work and btw the plugins should support Folia(just a tip) the plugins should also be optimized and use exploit fixer, cause u can crash a Folia server very easily;)
i am an expert; I’m just asking who is more experienced than me
what is that exploit fixer?
Also curious about it here. Should be something to add on the plugins I made, to help protect the server?
You really don't need an "exploit fixer" plugin. If there are actual bugs in Folia they should be reported to be patched, but if there's specific stuff then you'd want to patch your own fork. The "exploit fixer" plugins cause a lot more problems than they'd even remotley solve, especially these days.
i think such a major exploit/problem would be fixed ASAP or at least have a disclaimer on the website
honestly most of the things they "fix" dont really look like folia specific stuff.
this would imply that paper on its own is super unstable but it obviously isnt.
Such plugin is never needed on Paper or even Folia if you just keep your server updated.
they talk about "preventing known and private exploits" which seems suuuuper sketchy to me
Yeah, they are just trying to scare
Yeah exactly, I wouldnt trust such wording, it felt like those late night tv show advertisment lol
One way to know is to buy it and see the code
i mean flexing with "preventing private exploits" would make me think that the developer is the same one who developed these "private" exploits in the first place
Or just report it
Right lol
Crazy enough this is actaully the case for a lot of anti-cheat developers (who also make cheat clients...)
The best business is to sell to both sides
if you need money sure, its easy to prevent exploits if you know the implementatin really well.
but it is a sign for me to not trust these developers
never listen to people who want to sell you the solution to a problem they caused in the first place
well, if u think that, but i created my own mod to crash servers and it works on paper and folia so...
You report it
then why would you use a sketchy bandaid solution instead of getting it properly fixed?
Lisnia?
why would they report the issues and get it properly fixed if they make money of it?
developer of the sketchy exploit fixer plugin
alright let's stop this discussion, if you are not going to provide the exploit you claimed to know and still want to push for snakeoil solutions, I will take that you are just here advertising for it.
see thats the thing why report if u make money out of it
so until a report shows up in the #paper-exploit-report , I will ask you to not claim that you know exploit that's not fixed on Paper.
or I will be timing you out.
@tall thicket ^
nice ragebait.
ima end this conversation just by saying that this is a community project and community projects make stuff for everyone to enjoy without money being a factor
He literally just told us he's willingly to be a bad actor and that says a lot. Let's not engage any more with people who shown such behavior.
That is just being a jerk
just a quick question when will folia get further updates like 1.21.10 or 1.21.11? or will it even get one ?
No ETA 😔✌️
I hope it will get updated soon
is there any way to contribute to this project?
I dont think so
It is opened sourced tho
ima just fork ts and make my own folia :o
Eh good luck with that
thank you
Contribute to Ruthenium :C
whats that
wat if all the downsides of running Folia without the benefit of having a stable plugin API 👀 (interesting project though)
its a fabric mod 😭
Please don’t come to help channels to shill for your own software.
message.txt by @quasi imp: https://pastes.dev/zg817Lp5Ge
is this fine?
Hello. When will 1.21.11 be folia?
When will Folia 1.21.10 be released?
when someone updated it
Yeah, sure, but somehow Folia is just lagging behind, which is annoying.
well either don't use it or update it then if your business relies on it?
ye but theres basically nothing we can do 🤷
that's life when relying on free open source software
It's not my job to update PaperMC's software just because 19 people can't manage it.
entitled much?
feel free to not use the software?
if you profit off of the open source work it IS indeed your job to contribute to it's development lol
That's right, and I'm doing that by saying that Folia is lagging behind and needs an update for newer versions.
sorry, I meant "contribute positively"
That's positive because many could benefit from it, not just me, and just because someone lays the facts on the table and says that Folia is dead, you don't have to be disrespectful or feel superior.
how is being annoying about something not being updated yet positive?
do you really think the people that worked on the projects do not know the current state of it?
I am simply pointing this out and noting that there may be people without programming knowledge who are waiting for an update. This has nothing to do with me believing that the people who develop this software do not know what the status is. These are two completely different things.
people without programming knowledge should stick to Paper rather than Folia
Thank you for your constructive response.
I mean, even if you can't contribute to the actual work personally you can still contribute by paying devs for working on the software that you use. Investing into the open source projects you use should be part of your business plan...
Ah - we don't support offline mode servers. Sorry.
its test server
You have a login plugin and SkinRestorer. That isn't a test server. We don't support offline mode servers.
As I said, we don't support offline mode servers. Stripping out the questionable plugins and starting the server just to get support isn't going to change that.
the issue was the config
threaded-regions: grid-exponent: 2 threads: -1
In case anyone else has the same issue, when starting with folia
We should use a bot that responds directly 
If you want it to be faster you can donate to the paper project, or even fork it and make a pull request.
Really should lol\
Could make it scan messages for phrases like "ETA" and release data and just automatically send that lol
Just wondering, are there any problems with updating to 1.21.9? Or the developer just vanished?
The dev has still been working on stuff
folia is downstream to a whole bunch of other work
The dev has to I think make changes to concurrentutil and moonrise first
Hello! When will Folia be available on 1.21.11+?
🙁
The paper stable build hasn't even been released yet
hi anywhere I can like get archive folia 1.21.1 builds? I need that specific version for a specific datapack to test
I'd test with Paper instead.
You'd have to build 1.21.1
(And you shouldn't target 1.21.1 Folia at all)
I am trying to run jjthunder to the max
and that requires a very beefy CPU if you are doing single-thread, so I wwnated multi-threading
Chunk gen is already async on paper, and runs stuff in parallel where it can
Folia doesn’t support data pack command functions, idk if they use them but I’d imagine they’re potentially doing stuff that doesn’t work on folia
deploy it to prod immediately
😃
userdev release already out?
or do i have to publish it myself?
there is one publishing pipeline
outside of any odd circumstances if it's on fill it's on maven
eventually the maven build numbers will reflect that
Thats nice
Fill .. you are late buddy
How unstable are these builds?
Channel: Alpha
i would only use them to test during plugin development (but not final testing, i would test again once stable comes out).
just dont use them with the intention of players to actually play the game
So, not recommended
back things up and test it
How many players do you think a Folia server can handle on a Ryzen 9 9950X? (SMP server, where players will be building farms and machines, etc...)
with compression = -1
how spread are they?
Usually on most SMPs players aren't that spread, which is what you need for Folia
see ^
For an average smp with close together players it is probably better to use one of the following:
-Paper
-Leaf
-Fabric w/ performance mods
Like DonutSMP
Spread multi node
I was thinking fit about uh
500-700 on 7950x/9950x
big SMP network to handle 2000-3000
😐
Shard it
Folia isn’t intended for multi node use cases
Unless on each node they are also very well spread apart
yea i am
Im making like multi node exactly like donutsmp doing
im just wondering how much player i can fit until it reach 80-90% cpu
It’s going to depend a lot
With compression-1 , prechunk and optimized plugin
Well def plugin not gonna the one for this cause I have an dev gonna rebuild all the code to optimize
For reference my server gets about 40% cpu running fabric with about 40 ppl
On ryzen 7 7700
Which is a lot weaker
Paper should be similar
We dont talk about fabric 😭
Leaf is more optimized than paper
Ik im just using it for reference
For your use case with that many people fabric isn’t stable enough
But it performs close to paper
If you don't have a server yet, don't worry about running 200 (or more) players. Paper is going to be fine to start. As you grow you can adjust your setup to your needs.
Not every server type is ideal for Folia, but most servers just starting aren't going to want to spend the money for the hardware for Folia.
Getting 500 players takes like . . . years of a server being active.
Years and years.
Getting 20 concurrent players is incredibly impressive.
I already have an data before with 400-450 players using 80% of whole 7950x
My server network total last summer reached 2100
Im planning scale it to 5k
Okay then you already have a rough idea of how many players you're going to hold on given hardware - you can answer those questions far better than we can.
That doesn’t mean that much because they aren’t concurrent
It depends a ton on the setup, the server, etc.
welp im thought i can find some data in here
Ok so how many players does your server have online right now
At this moment
That will help us to judge
Again, it's not an easy thing we can throw out.
One server with a particular play style and setup with plugins could run 100 players.
Another server with another setup could run 50.
Bc like total joined doesn’t mean much…
You can't just divide a CPU by number of players and get math. There's a ton of other factors.
no no 2100 at same time last summer
not 2100 total
One poorly written plugin can slow down an entire server.
Is it a whole network? Or like a single survival world
I can tell, one plugin mess the folia thread = whole server cooked
whole network is 2100
Smp alone 1600
Same with Paper, yeah.
With 1600 just shard it
that what im doing lol
You will struggle if they aren’t spread really well on folia
to handle 1600
And if there is ever some event or smth and a lot go to one area ur screwed
But ocelotpotpie is right there’s no right answer here really
Esp. not with more information about how they are spread apart and like what kind of players are these (building large machines?)
farming machine
pickle
So they only build pickle farms????
wtf kind of smp is this
Good stats, you gonna tell Pedro to share the msot recent stat with me 
Yeah our goal is to inform users to the best of our ability, and I think making up number is going against that idea. 
haha
Can’t just throw out some number and act like it’s correct
Because we have no way to really know
😌 i will come back in feb or summer to share some datas
Please do! We love more data.
def i will 👌
@thin compass Would love it if you write a blogpost on your journey. Pedro told me he was gonan do it but he got busy 
idk about blog, but i will try share as much data i can
that would be nice. 
to able to get 5k players i need to start clip farming on youtube and tiktok 🙏
this year summer,i turn my server from 1k to 2k player under 1 week
lmao
6-7m views from tiktok alone bring me 1k players
prob i'm the biggest donutsmp fan lmao, been play like since 2023 and try to make it
also what else database stack do you guys recommend to use?
I'm using Redis and MariaDB right now
Again, it depends on what you're doing with it.
when regions are merged can they never become smaller again?
Not until restart, no.
uhh
