#waterfall-help
1 messages · Page 32 of 1
is that in the waterfall.yml file? let me try that
How many players can one Waterfall / BungeeCord instance hold? Has anyone had practical experience with this, or just theoretical insight?
after setting that option to true and restarting the proxy it's still giving the same error
actually hang on
nope, bad packet id
and the timeouts haven't been a problem without using proxy in the past
totokaka: afaik people usually look into load balancing around 1000
A thousand players on one proxy?
taht's assuming you don't have plugins on it doing dumb shit
Ok, thanks. Do you think it would matter what kind of load it is? Imagine everyone being in adventure mode, for example
well I guess it kinda depends on how much data you send through it (e.g. lots of new chunks vs. minigames with very few)
That does not sound like something you want if you ask me. What if the proxy crashes? Then you just lost a thousand players that want to login all at once.
well I didn't say that you shouldn't have a backup or any infrastructure around it
also if your proxy crashes then you have bigger problems imo
You can't really expect your software not to crash if you ask me, it happens to everyone at some point. Whether that is because one of your physical servers dies, because of a software mistake it will pretty much happen some day.
but you should also not waste resources just because of the fear of a crash. just make sure you have a backup proxy up and running ¯_(ツ)_/¯
is there somewhere else I should be setting timeout limits in waterfall? I have set them to even insanely high numbers and doesn't make any difference. -Dfml.readTimeouts are both also set very high on individual server and client. hasn't been a problem with other servers that had more mods
there is a 5 second timeout in what I think it waterfall.yml for the initial connection, but, no, the actual read handlers are stuck on 30 seconds
Ah OK thanks. So it’s a hard coded limit? Would there be any way to get around this without editing the actual code?
nope
Oh, tell a lie
It exclusively uses the timeout in the configs
ah, the connect timeout is a different config
which config? as i have looked through the configs and set all the timeouts i found to very high values and doesn't seem to have any affect
Well, if setting the timeout super high doesn't do anything, it's probably not that
just called "timeout"
yes that's the setting i have changed in the config.yml
You'd need sponge or something to be able to support bungee
But, the connection throttle is a CB thing, not a vanilla thing, so not needed there afaik
No idea if they added their own connection throttle or not
bungee in general modifies the handshake to be able to pass through the uuid/ip of the player properly
Is there a events hosting plugin for waterfall? Like a joinevent command and run cmds on event start maybe event create or gui I can set with that command?
any idea why my bungee instance running in Pterodactyl cant access the internet?
java.io.IOException: Could not get connection from URLs provided.```
is it because my daemon port is 8080?
isn't port 8080 used for HTTP ?
https://github.com/PaperMC/Waterfall/issues/418 probably related
@timid solstice
I don't know if there is a certain format to follow when posting for support or if this is even the right channel to post to. idk where else to go with this as the dev of the scoreboard plugin hasn't replied to anything and seems to not care. I've been having this problem with what seems as featherboard and bungeecord/waterfall. I've tired everything i can do to get it to work including removing scoreboard.data but basically all of a sudden when i started reloading my scoreboard plugin it would time me out on my client side. the server is 1.12.2 and i have tried everything. im more than sure its a bug with the scoreboard but i cant confirm 100%. So if anyone can help id appreciate it -thanks
https://paste.helpch.at/notezigewi.bash
oh and the error above is what i get when i reload the plugin. which is the exact error i get on my survival server. as this is happening on both servers of mine, but the error only post in the waterfall console
Don’t reload. Reloading breaks a lot of things. The system isn’t meant to support reloading

But this seems like there is a packet sent that already has been sent
i dont use /reload, i do /fb reload with is the reload for featherboard. im testing all my plugins again one by one as it seems like it has gone away by completely removing all plugins. so im adding them one by one back in and stopping and starting the server
@next nexus this error will continue happening that much is for sure
because the reload does not unregister the scoreboard team
and then attempts to re-register it which will make it error
because it still exists
hmm do you use an anti-collide plugin too?
yes i do, but it has something to do with protocolLib i removed it and it stopped
let me confirm it with my other server also
i think the last time i was testing each plugin i over looked it as i never had a problem with it before
yupp definitely something with protocolLib
Hi, is Waterfall pretty much a drop in replacement for bungee?
Anyone able to help? I just swapped from bungee to waterfall and I keep getting a connection Lost - Timed out after I join the server. It doesn't show me trying to connect to any of my paper servers and the logs in the bungee server look nominal
>.... [20:45:36 WARN]: Event PreLoginEvent(cancelled=false, cancelReasonComponents=null, connection=[/ip:51686|hhhapz] <-> InitialHandler) took 590ms to process!
>.... [20:45:37 INFO]: [hhhapz|/ip:51686] <-> ServerConnector [10b10t] has connected
>.... [20:45:37 INFO]: [hhhapz|/ip:51686] <-> ServerConnector [10b10t] has disconnected
>.... [20:46:07 INFO]: [/ip:51686|hhhapz] -> UpstreamBridge has disconnected
hey when i try to strt my water fall after config it theis happpens
on the bungee cord its also the same
how can i fix this
help pls
helloo
repair your configuration
Please paste large logs to a pastebin: https://paste.gg
is there a tutorial on how to use this ?
that generally means that you're connecting directly to the server and not the proxy
Any idea why my servers are only showing 4 columns of players in the tablist? I've got 90 players on one of them atm and tab_size is set to 200 in waterfalls config but it only shows 4 full columns of players so some aren't showing up.
Is 80 players just the hard limit for the tablist?
something like that yes iirc
Pretty sure you could get that functionality with something like LiteBans
Anyone know if I can add servers to the proxy without having to restart?
yes there's an api for that
Cool 🙂
That moment you cancel a ServerConnectEvent and wonder for 15 minutes why the PlayerQuitEvent isn't being fired 🤦
Are there any tests/read-ups on if network-compression-threshold for waterfall is better at 512 compared to the default 256?
it's situational
There is no magical "best value" here, if there was, am sure we'd be preaching it
do 5 bungee instead of 1 should decrease network usage between servers?
for example one waterfall use 700MB/s with 600~ players. but when i put a few more bg on same machine it will be same usage network, am right?
the backend servers will see the same traffic
if you're running the proxies on the same machine, same situation too...
[00:08:29 ERROR]: SLF4J: Found binding in [jar:file:/home/container/plugins/altfinder-bungee-3.4.24.jar!/org/slf4j/impl/StaticLoggerBinder.class]
[00:08:29 ERROR]: SLF4J: Found binding in [jar:file:/home/container/waterfall-341.jar!/org/slf4j/impl/StaticLoggerBinder.class]
[00:08:29 ERROR]: SLF4J: See http://www.slf4j.org/codes.html#multiple_bindings for an explanation.
[00:08:29 ERROR]: SLF4J: Actual binding is of type [ch.qos.logback.classic.util.ContextSelectorStaticBinder]```
Anyone know how to resolve this? Or if it even matters?
It's a problem from the plugin
for example one waterfall use 700MB/s with 600~ players. but when i put a few more bg on same machine it will be same usage network, am right?
@tidal musk i think you mean 700Mbit/s, but its too much for 600 online, it should be ~150-170Mbit/s
Its not my screen, but who made it says that this is stats for 1300 online
is it bad?
bruh I use BMON
are there also verification issues with waterfall?
that i need some plugin or so?
how do i make it so when a server closes it redirects people in a different fallback server
instead of logging them out ?
use a plugin
any recommendations ?
Project Waterfall build #342: SUCCESS in 3 min 15 sec: https://papermc.io/ci/job/Waterfall/342/
Project Travertine build #138: SUCCESS in 1 min 35 sec: https://papermc.io/ci/job/Travertine/138/
I'm using waterfall and I can't connect to one of my servers. It says the default or fallback server cant connect. My other server works fine and I changed the config. Any ideas?
Bad configuration
I did the same things on each of em @kind karma
what did you do?
plugins with bungee-mode will also work with waterfall?
@kind karma any ideas ^^
did you configure waterfall properly?
What's the difference between having UDP Query on or off in bungeecord (query_enabled: true)
that protocol exposes more information
you have an invalid character in the modules.yml file
just nuke it
it'll be regenerated
however if that happened out of nowhere
i strongly suggest you to run disk check
let's see... fsck
if it's your own machine
linux?
oic
okay
yeah i cannot help you with windows, perhaps someone else can
inb4 cryptolocker
"we yoinked your modules.yml file, pay 1btc to unlock"
it's not important
for overall functionality
it simply tells your proxy where to grab modules for /server /alert etc.
nah
where it'll download them when they don't exist
you can download them yourself freely
So im about to set up a network for our community,
We currently have a paper survival server, should i go waterfall?
I see lots of plugin that support bungee but does that also work with waterfall?
correct
Hello, I added a modded server to an existing waterfall setup, and the modpack itself is pretty big. When connecting, some slow clients fail to respond with alive packets, and it gets them disconnected with a read timeout. I've made sure that the server itself has the read timeout set to 5 minutes, using the RandomPatches mod, and the timeout in config.yml of waterfall is set to 120000. Is there anything else I can do?
let clients increase ping timeout as well maybe?
-Dfml.readTimeout=600
or w.e it was
not sure if it worked on client though
but since i assume it's forge then it maaaybe works
.g fml.readTimeout
(DiscordBot) https://www.curseforge.com/minecraft/mc-mods/randompatches -- RandomPatches - Mods - Minecraft - CurseForge: "The login timeout generally has more impact than the read timeout. This also sets the "fml.readTimeout" and "fml.loginTimeout" system properties."
Looks like the loginTimeout is set to 900 in randompaches, as well well as the readTimeout to 600 (server-side). In the console of waterfall, there is this:
If this is the case, isn't it the server (waterfall) closing the connection?
On the server itself, the client is still in a connecting state, and waterfall intnerrupts the process
All other clients work fine, it's just the slow ones that do this
The server says the client has just disconnected, it does not complain about timeouts
I thought the timeout setting in config.yml of waterfall is the source of the problem, but that is set to 120000, and people say they get disconnected after about 60 to 100 seconds
Alright, will reloading with greload work for crude testing without restarting the proxy?
shoould work
well according to the code it should in theory
timeout value is definitely respected even in ServerConnector
checked it right now
(DiscordBot) https://github.com/SpigotMC/BungeeCord -- SpigotMC/BungeeCord: BungeeCord, the 6th in a ... - GitHub: "BungeeCord, the 6th in a generation of server portal suites. Efficiently proxies and maintains connections and transport between multiple Minecraft..."
very handy this bot, checking
oh, alright, so it's the right timeout to use
that's the channel initializer bungee uses for all the channels it needs to initialize
which means client<->proxy and server<->proxy connections will both use same timeout value
not sure what's setSoTimeout is
but timeout is handled by netty's thingy
.g netty ReadTimeoutHandler
(DiscordBot) https://netty.io/4.0/api/io/netty/handler/timeout/ReadTimeoutHandler.html -- ReadTimeoutHandler (Netty API Reference (4.0.56.Final)): "addLast("readTimeoutHandler", new ReadTimeoutHandler (30); channel.pipeline().addLast("myHandler", new MyHandler()); } } // Handler should handle..."
Alright, well that seems to have improved things. Now I'm getting this:
[13:31:29 INFO]: [MAXX96|/f:53256] <-> ServerConnector [Lobby8] has connected
[13:31:42 INFO]: [MAXX96|/f:53256] <-> ServerConnector [FTBR] has connected
[13:32:41 INFO]: [MAXX96] disconnected with: [Proxy] Lost connection to server.
[13:32:41 INFO]: [/f:53256|MAXX96] <-> DownstreamBridge <-> [Lobby8] has disconnected
[13:32:41 INFO]: [/f:53256|MAXX96] -> UpstreamBridge has disconnected```
now server closes the connection
This is from the server log:
[14:23:57] [Netty Epoll Server IO #3/INFO] [journeymap]: CLIENT
[14:23:57] [Netty Epoll Server IO #3/INFO] [FML]: Attempting connection with missing mods [spongeapi, sponge, spongeforge, bedpatch, creject, fastasyncworldedit, worldedit, prism, spark, tellme, nucleus, placeho
lderapi, universeguard, mmcrestrict, luckperms, catclearlag, mmcreboot, redprotect] at CLIENT
[14:23:57] [Netty Epoll Server IO #3/INFO] [CodeChickenLib-ConfigSync]: Skipping config sync, No mods have registered a syncable config.
[14:25:57] [Server thread/INFO] [minecraft/NetHandlerPlayServer]: MAXX96 lost connection: Disconnected
It does not look like it's the server closing the conenction to me, it's just like before, when waterfall was timing out
I should try restarting waterfall just in case
i mean if proxy says it lost connection to the server
then server closed it
but yeah never ending nightmares with proxying modded servers
I expected it to be a nightmare, but this is way over my expectations, lol
i had to do that once, it worked pretty well however
there's also Velocity proxy
which might work better than waterfall or any other bungee fork
worth a try, at the cost of not being able to use bungee plugins
The timeout in config.yml is now set to 360000, so that's 6 minutes, and it still happens after 60 seconds. It might be something on the server side, I'll do some more server-side research
I have to stick to waterfall, because the existing server setup is based upon it
is a plugin which prevents the login without mojang verification needed?
like if they were to connect directly to a gamemode-server instead of the waterfall
You could use some iptables rules to only allow connections from the waterfall server to the gamemode-server, or an ip whitelisting plugin with the same principle
bungeeguard
fuck ip whitelist plugins, they'll either not work or can be worked around (latter in shared hosts)
Good to know, I'm using iptables to only allow connections from the waterfall server, hope that's safe
.g lucko bungeeguard
(DiscordBot) https://github.com/lucko/BungeeGuard -- lucko/BungeeGuard: A simple plugin that verifies whether...: "BungeeGuard is a pair of plugins which intercept the BungeeCord handshake protocol, and allow backend servers to verify whether players connected..."
[Netty Worker IO Thread #2/INFO]: [Storm974] disconnected with: ReadTimeoutException : null
fix? 😢 many players got kicked .
that means that the proxy didn't recieve any packets from the client in so long
[Pterodactyl Daemon] Checking size of server data directory...
[Pterodactyl Daemon] Disk Usage: 54M / 100000M
[Pterodactyl Daemon] Ensuring correct ownership of files.
[Pterodactyl Daemon] Running server preflight.
[Pterodactyl Daemon] Starting server container.
[Pterodactyl Daemon] Server marked as OFF
[Pterodactyl Daemon] A fatal error was encountered while starting this server.```
wow 😦
guys i need help please with bungecord
Hello, any idea for this (just look upper )
Yes, but, really, use a permission plugin
Hey,
i changed the ports of my Servers and now i cant connect to the fallback?
https://pastebin.com/kDcMumj5
https://pastebin.com/BGUigzW8
https://pastebin.com/uHJ0AWKv
Can you help me?
The server at that time received a flood or a fake session attacks
still a shitty plugin which blew up
ok
and my problem? 😄
connection timed out means just that, the connection timed out
generally down to network config
how to fix?
Quite simply, work out what's wrong with the network
but i dont get it xD
I've been getting lots of kicks
Regarding the scoreboard
[18:17:49 ERROR]: [/redacted:13369|redacted] <-> DownstreamBridge <-> [factions] - encountered exception
java.lang.IllegalArgumentException: Team 31c3a075-cd10-4 already exists in this scoreboard
at com.google.common.base.Preconditions.checkArgument(Preconditions.java:191) ~[waterfall.jar:git:Waterfall-Bootstrap:1.15-SNAPSHOT:03b43fd:unknown]
at net.md_5.bungee.api.score.Scoreboard.addTeam(Scoreboard.java:73) ~[waterfall.jar:git:Waterfall-Bootstrap:1.15-SNAPSHOT:03b43fd:unknown]
at net.md_5.bungee.connection.DownstreamBridge.handle(DownstreamBridge.java:231) ~[waterfall.jar:git:Waterfall-Bootstrap:1.15-SNAPSHOT:03b43fd:unknown]
at net.md_5.bungee.protocol.packet.Team.handle(Team.java:124) ~[waterfall.jar:git:Waterfall-Bootstrap:1.15-SNAPSHOT:03b43fd:unknown]
at net.md_5.bungee.netty.HandlerBoss.channelRead(HandlerBoss.java:105) ~[waterfall.jar:git:Waterfall-Bootstrap:1.15-SNAPSHOT:03b43fd:unknown]
This is the amount of kicks in the past month with this issue. It started on one server and it skyrocketed and began to happen on all of our servers
basically, bungee doesn't like it when you move to a server which has a scoreboard with a team the player already has on the current server
the weird thing is that not all the servers use the same scoreboard pluign
plugins basically need to use a mechanism to avoid that, looks like the plugin is being dum and using the players UUID
We have like 4 different ones (on different servers)
@reef fulcrum That plugin you are talking about is featherboard and I believe that it generates a random UUID to precisely avoid this issue
Let me check our changelog to see what we updated on April 13
It seems we updated to Paper #178
You're like 100 builds behind
Might want to look into updating again
A lot of stuff had changed
You are not following what I'm saying
The curve of players getting kicked started rising on April 12, when we updated to Build #178
We are now on build #260 on all of our servers
Ahh
At this point I'm unsure if it is related to paper or waterfall
It spiked again yesterday
After this commit
@unique linden
you said it spiked in april, well before that commit, i dont think its related.
point being its something else causing it, the source of the spike is unknown, and when it happens is inconsistent
we need to just try to improve waterfalls team shit though
the team thingy?
or papers team name, could use timestamp in it
just make it not blow
Imma just override the downstream bridge and ignore the errors for now
hey aikar, how much of a performance increase would I see switching from bungeecord to waterfall with 150-200 players?
.try
definitely worth switching, you won't lose anything
Only use of a team in paper is for the collide rule as far as I’m aware
@tidal musk that page is amazing
It already has a random string appended to the end for this reason
I'll try out Firefly
sure thing, however no support with that in this guild
why arent some of these PR'd to waterfall @tidal musk , whats pros and cons of this TCP Fast Open
honestly
if you look at the pr queue sometimes
then shit takes ages
can't blame cat either because of his health issues really
but it is what it is
sorta same reason why leaf started his fork
shit simply takes ages
Y'all need another contributor for waterfall?
Or is the issue lying in getting PRs merged?
Some of the PRs are honestly dum/confusion
https://github.com/PaperMC/Waterfall/pull/395/files seems reasonable?
what does bungee do to tablist rewrint regarding https://github.com/PaperMC/Waterfall/pull/386/files
that one I was conflicted on
But, actually, looking at it, this is just... o.O
Only really relevant for online proxy offline backends
offline UUID > online UUID mapping
so its only useful for me and xdark lmao
wait no papers bungee=true also uses online uuids right
correct
so that seems reasonable to disable if using paper
or if could detect it somehow and just auto
I don't really get why you'd ever be running a server like that
with bungee=true? everyone hopefully does
All that option really saves us is a hashmap lookup and contructing a new packet
or do you mean worthwhile of the feature
It's in the guide to set bungee mode to true.
it does seem like a micro opt
does anyone even run into performance issues on the proxy?
i sure as hell dont
cpu 15%
The thing I have is that I'm dead, and some of these I did kinda misjudge what they where doing as my head is 10/10, some of them where just pretty o.O at the time
Oh there's only a few PRs in the backlog. I was under the impression it was as much as Paper.
i guess those with 100+ concurrent could maybe hit more usage
as i have multiple proxies so my loads distributed
I think people usually say like 512MB per 50 players
Doesn't seem to be super intensive
Stuff like that one, I can see some of the gains, it's just a pretty bleh of a patch in terms of how invasive it is
i have few safe to pull patches tbh
like using asm to execute event handlers
and perhaps even using enums for scoreboard stuff
latter patch is very safe unless mojang is going to change scoreboards
or there are dumb plugins which do something very hacky
that's safe to pull
similar thing exists in paper
also safe
latter will give more reasons for people to use waterfall instead of bungee
https://github.com/mikroskeemsrealm/Firefly/blob/master/Waterfall-Proxy-Patches/0031-Save-memory-on-scoreboard-teams.patch is the scoreboard thingy i mentioned
sure, can make prs if interested @unique linden
but md5 said the asm patch is pointless!
MAH NANOS
yo i know it's a newbie question at first but hear me out, i reinstalled the latest update of waterfall and when i use the commands such as server or ip or other proxy command, it says you do not have permission to execute the command even tho i put my username on the group,..
that intern one is dangerous
i wouldnt do that
your more so wasting memory
with papers random team names, your gonna fill up the intern pool with a lot of garbage
yeah
that was my concern as well
i still did that
well because it's my experimental proxy fork in the end
but dropping the string intern shit
what do you think about using enums there
but md5 said the asm patch is pointless!
markro mirco opts add up ;)
enum seems ok and i guess you didnt intern team names, but player names and such too seems same issue lol
basically in my situation intern kinda worked ok; because i mostly had duplicate team prefixes/suffixes tbh
I was actually considering about dropping that, but pretty sure I dropped some other collection that was hitting perf and people had a blow up in memory usage...
:L
i used to set up rank prefixes mainly
even if you have 300 players with a 100 char prefix, round up to 30k of memory... thats nothing for "live data", let GC clear it when its done, but interning will make it sit around after the player logs off
yeah intern is completely useless in that case
but people using just color codes
I think that it was more written around the times when featherboard was being shit
Well, super shit
i.e. 100mb scoreboard.dat files
interning shouldnt of solved that though
also interning has a cost to invoke. i ended up removing it from timings
at least i think i did
yeah i did
Lemmie remember how to git, I was already in minds of dropping that one and well, that double confirms it's death
i have a little question can i ? xD
yo i know it's a newbie question at first but hear me out, i reinstalled the latest update of waterfall and when i use the commands such as server or ip or other proxy command, it says you do not have permission to execute the command even tho i put my username on the group,..
use decent permissions solution on bungee
such as LuckPerms
save yourself from the endless agony you're going to face with bungee's miniature implementation
Project Travertine build #139: SUCCESS in 1 min 58 sec: https://papermc.io/ci/job/Travertine/139/
ye
Project Waterfall build #343: SUCCESS in 3 min 25 sec: https://papermc.io/ci/job/Waterfall/343/
cat id really love to see WF be the first class for most people and demote travertine to 1.8 status. i personally dont care for extra code dealing with 1.7, but would want snapshot protocol support
Issue with the proto stuff is that it's literally short lived and often makes merging with travertine "fun", which makes for logistics fun, e.g. many of the people wanting that stuff are already running trav because 1.7
Which is somewhat a irk given that the only real reason I used to throw that stuff into travertine, especially as I generally just force disable entity metadata remapping for those versions vs actually dealing with that mess now
People apparently do :L
i dont
you don't yes
or at least just make bug fixes
let them boomers use an outdated version
honestly dropping trav with 1.16 release would be ok
if they gonna run 1.8 servers they dont need updated shit
imo
+1
other community members can probably run a trav fork themselves fine
or could let someone in 1.8 community maintain it lol
ye
and be "not our problem"
I like the way you think.
i dont see anything in the 1.7 patch that makes adding futrue protocols easier, its the PATTERN that would be duplicated (if needed)
for pvp is it good to use for example a 1.15.2 server with anticooldown plugin that is updated or using a native 1.8 server that is outdated ? is there differences ?
it's just a general question
thats paper question, this is wF
? wF ?
oh sorry
oupsi
about my problem again, markroskeem when u said to use luckperms, i already have it and i need to put bungeecord perms to luckperms is it ?
both
Yea, you'd need to manage your bungee perms on LPB
when i put the bungee perms in luckperms, it still says you don't have permission
are you using /lpb or have it all networked against one mysql server?
i setup the luckperms on bungee and it still doesn't work
defn bungeeguard
great thank you
i opped myself but i can't even talk in my lobby
You do not have permissions to perform this action!
i resolved my problem regarding luckperms, thx guys 😉 luv u
I'm getting this error during a bot attack. Proxy has gone down twice in the past hour. Any thoughts? Individual errors for each player/bot who joins. Command spam to crash it? Not quite sure. Urgent help would be very appreciated :)
.try
sounds like they havent been setup correctly then
anyone knows how many memory allocated needs waterfall for a server of about 100 users?
Hello! I am working on a one time live event in Minecraft. We are expecting a few thousand players and are estimating 1000+ players per waterfall instance. We are running this in the great cloud with a loadblancer in front.
Is there anything we can/should do to improve waterfalls performance?
Any recomendation for hardware needed for 1000+ players per Waterfall?
We have one node for each Waterfall instance
another music festival?
Norway's national day celebration 🎉 🇳🇴
ah
a tutorial says that i need to set bungeecord to true\
for my servers
but i only have bungee-online-mode: true
is thatt the same?
no, you want spigot.yml
Hmm, probably a general bungeecord question to be fair, does waterfall/bungee sent a new Game Join packet every time the player changes server?
waterfall does if entity meta rewrite is disabled
which is not default
bungee doesn't support toggling that, it rewrites entity ids and plays with world change/respawn packets
so usually it will never send a Game Join packet
So atleast I know that I need to be a bit careful tracking the client's own world
alternative proxy, velocity iirc sends join game packet by default
Waterfall or travertine fix Netty Exploits ?
Anti Xray does not seem to be working
Is there some kind of selective online mode? I am trying to allow a local ip connect in offline mode for a bedrock proxy but refuse cracked java edition clients.
maybe via a plugin, but not natively
Ok time to ask that in another channel then :)
Project Waterfall build #344: SUCCESS in 2 min 59 sec: https://papermc.io/ci/job/Waterfall/344/
travertine should auto update ; ;
Project Travertine build #140: SUCCESS in 1 min 21 sec: https://papermc.io/ci/job/Travertine/140/
is zzzCat the infamous electronically charged feline or some other cat?
he is one of the strays
he comes around sometimes when he's too tired to hunt mice
oh god he has spawns?
pets cat
oh but he's not spying for sparky
ok
just trying to learn what cat's who
the charged one was up past 5am, got a little grumpy there toward the end, thought maybe he changed his name to zzzCat for a time
then zzzCat is safe to pet? pets zzzCat
yay i still have my hand!
see it's a good thing i checked
I am new to the waterfall / paper world and I have problems getting my server to work can someone send me an example of an configured waterfall server template so I can see what I did wrong
thanks already
PS: the waterfall wiki sux pls update it
just follow the bungeecord one lol
not do you get help because you don't post what you've got 
K
The problem is that my server is online but if I try to join the server using the ip then it says that it's ofline
doesnt sound great, does it
now if you post what you've got, we might be able to help you
dosnt sound promising
xD

Beyond that, it's all just a typical bungee install, redoing the documentation for waterfall is on the todo list, but like... amded
Ok
Optimized & Recommended JVM Startup Flags: https://mcflags.emc.gs
@lunar rose see pinned messages
at what player count should one start to consider multiple proxies? 200-250?
It's really personal opinion, I've seen a single proxy handle up to 500 or so
That's also true
I've also done projects where I run 4 or 5 for 300 to round robin and location based.
Ahh that’s pretty neat
They will connect to the same backends but the proxies just are spread out a little. US, EU, CA, etc.
To try and give players the fastest connection to a proxy.
there any video tutorial anywhere thats in english for setting up waterfall and explaining the differences between it and bungeecord? only thing I've found so far is in spanish(or french? dunno, couldn't read the title so I skipped over them)
I handle up to 800-1000 players per proxy, it depends on what plugin you use and what is the power of the host
whats the help command for waterfall?
none
Hello,
Is it possible just to reduce ALL this big line and show "Waterfall" ?
Because it's ugly to see all theses informations for all players :c
Hey last night just everyone disconected and coudnt join anymore with timed out
https://pastebin.com/fbiNcDtm
hi, what is "game_version" in waterfall.yml?
Is there something so i can sudo a player to run a bungee command?
your servers are creating duplicate teamnames
it's basically a well known limitation of bungee that you can't jump between servers sharing team names
well
Really gotta look into just queueing those packets... bleh
Oh, so, a plugin on the server is being dum then
Might wanna try the customary "delete scoreboards.dat"
Yes, something is creating team packets improperly so it seems
The vanilla client will just scream the errors in the log file afaik, bungee is more sensitive about it, especially as it needs to track that stuff properly to work
I always wondered why is this a limitation ?
Why not just re-using the old Team object
Or recreating one
because it's invalid state
Also, that one specifically is generally triggered mid server switch, right before bungee clears the scoreboard data from the client
but i didn't update nor add nor delete any plugin and it just happen suddenly
best workaround I got for that is basically gonna be to just queue those scoreboard packets, which, am not fond of, but, not really sure what else can be done there
i didn't even want to change server, i just login and it kicked other people out with this error
Big issue sorta not really. When I join my proxy it takes about 2-4s to join. It stops at the encrypting screen turning this time
hi
anyone can help me?
if i connect 2 server with waterfall (proxy), it is necessary that the 2 servers got waterfall installed or with only 1 server that got's waterfall it works?
in 1 server waterfall and on the other paper or spigot ?
@timid solstice
It's a proxy, general setup is to just install it on one machine and configure it to connect to the servers running on the same/other machine
(DiscordBot) How To Make A BungeeCord Minecraft Server - How To Make A Minecraft Server Network Ep. 3 - length 38m 43s - 633 likes, 33 dislikes (95.0%) - 38,776 views - The Breakdown on 2018.09.13
Just, make sure that you use bungeeguard and/or read the firewall guide
ok, thanks
some chat message, maybe?
did that dude also make a waterfall setup tutorial?
I whitelisted myself on fallback
I try to join server
It says i am not whitelisted
what
Hi, I'm getting this error a lot when trying to connect to my server
java.net.UnknownHostException: failed to resolve 'sessionserver.mojang.com' after 2 queries
Anyone know?
Installing waterfall is it just a drop in replacement for BungeeCord?
yes
You're having DNS issues
You can disable netty DNS in waterfall.yml which will also make it use your hosts DNS setup
Beyond that, check firewall/network
is there a way to add groups rather than just individuals to the config for permissions
Really, use a permission plugin
LP is basically the go-to these days, can even be all networked up and all that
trying to avoid that as we are having box issues right now
can run it standalone too
but thank you for the quick response
You'd just manage all your permissions with the lpb command with it instead of lp
ok sweet thank you very much
anyone knows a way to see something like timings for waterfall? i have high cpu usage and i know it's not waterfall, but i wish to know what plugin causes this cpu leaks.
You need a profiler
ah
Big issue sorta not really. When I join my proxy it takes about 2-4s to join. It stops at the encrypting screen turning this time
Anyone know how to fix this issue?
Hi! Super quick and easy 3 questions if anyone has a sec!
- Does Paper remove any functionality from Vanilla Minecraft?
- If I switch to Paper and it doesn't fix the issue with my server like the guy at server.pro says, can I revert to vanilla easily?
- When a new version of minecraft is released, how long is it usually before a new version of paper is released? (Like more or less than 2 months?)
Thanks! I haven't heard of paper before today and I am curious!
copy that question into #paper-help =)
this is for waterfall which is something else
oh I am dumb, sorry!
https://paste.helpch.at/zohibumiga.bash
Getting this error. It is kicking players whenever someone joins a server
Running waterfall proxy with paper as my spigot server, but the error is on waterfall itsself
Question, Is there a plugin where a player can setup a sign shop and put on the sign somethin where you can buy 1 diamond for 20 gold or somethin
does waterfall work for sponge? if so then were would i get it from? tyvm for the help 🙂
@tidal musk i believe essentials signs has that feature and you can give players perms to each type of sign
it also supports trade signs where you load something in the sign that players can trade for and once it runs out it must be re-stocked, sometimes it can be more secure to give to players something that will run out than access to buy signs
please use correct channel #paper-help
i ued to use Leymooo bungee fork for antibot
its plugin or waterfall fork?
where u download it?
.g github leymooo
(DiscordBot) https://github.com/Leymooo -- Leymooo (Vjatšeslav Maspanov) · GitHub: "Sign up for your own profile on GitHub, the best place to host code, manage projects, and build software alongside 50 million developers. Sign up...."
how about anti exploit and null pinger?
isnt thta "kinda" fix on waterfall?
i mean if i use that Leymoo bungee fork, means i am not using waterfall anymore?
which has more less cpu usage? @cold notch
yea
The hecc is that
Hey
Im getting "ReadTimeoutException : Null" as an error when some(most) players try to join the server
how can i fix that every player has the same exact ip adress
i used both methods of ip foward
and im self hosting
how can I make like people can't join paper servers without going through waterfall proxy ?
Yes, you do want that
why ?
configure your firewall properly or use bungeeguard
@tidal musk in case you got minigames (just a case) and every minigame is hosted on a separate server you don't want players to join the server without going through lobby you first want to get a full party before they join it to make a queue for exemple.
If someone can join it directly it will be annoying because you have a 4 player party that will be included in a server where someone belong so there will be 5 people in a game of 4 it will go wrong
no reason to delete
np
accept your failure
;)
your messages are still on irc
i thought that was the totall standart having players through proxy
well if you set up your network
then you always want players through the proxy
yes i know
ok and btw can u help me with my problem
@reef fulcrum you meant BungeeSafeguard ? (the one that is waterfall version)
same IP means that you didn't configure ip forwarding/bungee
Don't ping, literally "bungeeguard"
i did
and i have this same problem with my vps that i use for mc server
it shows the same ip that every player has
is it possible to have servers that are not requiring proxy to log with bungeeguard ?
in my experience its unsafe
No, you can't connect directly to the backend servers
Basically, if you can see the servers IP forwarding warning, you have a security hole
that's for private use that's not a problem at the moment just got a server with friends that I want to have in bungee and without bungee
not possible, basically
Nor is there really much of a reason to do so
if you just want people to be able to go straight to a server, use a subdomain and throw it in forced_hosts
just because I want to test some functionnalities of bungee and don't want for my friends to have to come through it just because of tests if I can't that won't be a problem just wanted to test a lil' ^^
bungee in spigot.yml
is set to true
god knows, if both those are set properly, it will pull the IP/UUID from the proxy
ok what settings should i have then
best guess is stop the server/proxy and double check those settings, make sure that the files saved/uploaded/whatever properly
these are my setting and all paper servers have bungeecord set true
and this is the ip every player has
can someone explain me why am i getting this
instead of public ip of player
best guess is either not configured properly or some plugin is being weird
you can see the config
that's only the proxy config
what other config do u want to see
Which, if the server is also configured properly, rules out the first one
spigot.yml from your server
In that case, test without plugins
How can I get the sourceCode of Waterfall
github
zzzCat
im still waiting
@reef fulcrum im sorry but im waiting for an answear for and hour
who can help me your are the dev and i called isp they said all is ok
as I said, test without plugins
and i did
If you've configured it properly, it will be sending all the data to the server
if it's not, something with your setup is busted
@tidal musk its def something abt u self-hosting
I use waterfall for abt 4 servers and it works pretty much perfectly
you thinks i haven t spend 3 weeks fixing this
Maybe I could help ya, dm me
does waterfall use a lot of cpu or does paper?
are there bungee plugins that launches servers on event ?
May I ask if this is some error or so with Paper or LaggRemover and what it does
[15:43:47 ERROR]: Could not pass event AsyncPlayerChatEvent to LaggRemover v2.0.6
java.lang.IllegalStateException: Asynchronous getNearbyEntities!
at org.spigotmc.AsyncCatcher.catchOp(AsyncCatcher.java:15) ~[patched_1.15.2.jar:git-Paper-282]
at org.bukkit.craftbukkit.v1_15_R1.entity.CraftEntity.getNearbyEntities(CraftEntity.java:535) ~[patched_1.15.2.jar:git-Paper-282]
LaggRemover is calling a function async it shouldnt be
their problem
also wrong channel
oh right sorry my bad
be happy that AsyncCatcher caught that operation...
bungeecord.command.send
if you have luckperms installed on proxy
yes
if you don't
then no
How do I listen to ProxyQueryEvent, because this doesn't run?
@EventHandler
public void onQuery(ProxyQueryEvent e) {
System.out.println("debug");
}
Hello,
is it possible to make timings of waterfall to see why my waterfall uses 18% of CPU with 90 players ? ( I get a xeon e-2288g )
(Waterfall consume MORE CPU than my 1.15 server with 60 players online)
all server get the same CPU ( xeon e-2288g )
you could increase compression threshold to test it
you could increase compression threshold to test it
Which settings do you recommand for a bungee with 150 players ?
@tidal musk
Exception in thread "main" java.lang.IllegalArgumentException: Server hub-2 (priority 1) is not defined
rename it everywhere
or go back to hub-2
in all needed configs
you have a server in the priorities list which is not defined in the proxy
either re-point it to the server or remove it
Crap, ignore, I didn't scroll down herp derp
Does anyone have any idea?
https://discordapp.com/channels/289587909051416579/555462214606389275/709839034398146701
@topaz ingot You probably aren't pinging the proxy with the query protocol. You probably want ProxyPingEvent
I want to be able to sent a custom motd string, because I wanna make use of the new RGB colors from 20w17a
proxypingevent is right then
RGB colors are not supported at all yet. You'll need to wait for a 1.16 update for BungeeCord first.
I have a working 1.16 proxy and server, but I wasnt able to change the BaseComponent to add support for the new colors
will 1.16 support rgb colors?
...well, you need to change the BaseComponent to support the RGB colors.
We've started an enum of color names from https://colornames.org/
oh no
Right now we just have it fetching from https://colornames.org/fresh/json/ every 30 seconds and it will go through and create a color option from each entry.
We figure by the time that 1.16 comes out, we'll have a solid start on the colors that we can allow servers to use.
"Sir, do you have this shirt in the color 'Pp Yes'?"
i saw that too... second refresh
What are the first things you guys do when you want to start up a network and you have pterodactyl panel on your dedicated server? Is it for example adjust a important config that reduces lag with a lot without changing the gameplay or a startup command or anything else that I don’t do now?
Is BungeeGuard enough to protect your network? I have 2 nodes, the backend machine is not protected by any firewall for now
But there's bungeeguard setup
I’ve heard bungeeguard is better than IPwhitelist if you don’t want to mess with iptables
but iptables is still the best way!
@untold leaf are you the real techlead
because you use docker incorrectly

Hello!
I have an issue with MOTD. It’s exactly the same as on server.properties from the primary server.
IPtables doesnt work well with docker
absolutely shit claim
I have an issue with MOTD.
so this motd is copied into your waterfall config as-is?
Yep.
so you probably need to unescape the motd
The normal color code with the &?
the \uXXXX thingy
Gonna thry that. Thanks!
Worked fine! Thank you @tidal musk !
anyone know of a way to make the players online appear when you hover over the playercount in the server menu?
afaik you can modify that in the ServerListPingEvent
FF00;Should we add rgb color support like that? Or #22FF00 like that?
Hello,
I set in the config.yml :
network_compression_threshold: 1024
but my CPU usage of the bungee is still 15-20% with 100 players ( like when I got network_compression_threshold: 256)
Hi is bungeeGuard enough to protect your backend servers?
no
someone could still ddos it
properly firewall it
(if you mean "enough to protect from somone logging in with a fake player", sure)
Having difficulties setting it up with docker
@hallow echo bungeecord uses resources as well
but more than 2 1.15 server with 100 players
what is this warning? : No client connected for pending server!
check the pins! @tidal musk
https://hastebin.com/qikazumaqu.bash
Error in my waterfall/proxy console. Scoreboard error of some sort.
There is no error on the paperspigot servers.
Plugins:
https://hastebin.com/idohomagit.md
can removing the scoreboard.dat for my worlds cause issues?
they don't really contain data
depends on if you have a plugin or something scoring data in there
its just featherboard making stuff
if not, not really, any plugin should be capable of recreating its stuff
hey zzz do you know what the warning ‘no client connected for pending server’ is?
the client disconnected or was kicked before being connected to the server there where going to connect to
Ah thanks. Is that something with a plugin or just their connection to the proxy.
Is there any benefits to running a waterfall node under another waterfall node?
as in, having two proxies
no
FlameCord simply fixes Netty Exploits by following this procedures:
· Flush() before closing a connection. (Closing them normally seems to keep the connection instead of closing it)
· Close on invalid request. (If a invalid request is done FlameCord closes the connection)
· Close on exception. (If a exception occurs because invalid data is sent FlameCord will close the connection)
Simple as it sounds, this is the only procedures you need to prevent Netty Exploits.```
Is that something that could be done to waterfall as well?
The actual issue was fixed by netty already.
but still needs to fix thing, when packets still reading after channel close, to completely fix this issues.
This one is kinda netty fault since it still attempts to call decode without checking if the connection is closed.
But yeah, you can do a workaround by making the handler a single decode, and manually decoding the data while you can. That way throwing exception will also discard unread data.
Actually making a splitter work that way also increases the perfroamnce slightly due to less methods invocation.
.paste
Please paste large logs to a pastebin: https://paste.gg
please help
Does Travertine 1.15.2 work with normal bungeecord plugins? Or are there any changes in the API? There shouldn't be or?
Hello, does anyone know how to make the bungeecord kick message reflect spigot one? In spigot \n works but when I reconnect via the bungeecord network newlines are ignored
for waterfall, is there settings to set player location back to Lobby when logout or disconnect? without force default server? Because force default server cant count all player
micbaw: force-default-server is exactly what you are looking for. not sure what you mean by "cant count all player"
LoneDev: I'm pretty sure bungee doesn't forward kick messages by default? might be some plugin
use a plugin
my player is 150 right now
but it show 1
what plugin? and put on the waterfall or lobby?
it's a list of servers
It tries to connect in order
But sort from the top?
Example:
Priorities:
- Lobby
- Survival
- Minigames
all player in those server will be shown on server list? and will try to connect from Lobby?
uhmm
quite simply, you wanna leave force_default off, and use a plugin to throw them to the default
or, use force_default, and use a plugin that lets you control the motd, etc
i moved from flamecord to waterfall, but waterfall saved player location on disconnect, but its not good to me because login authentication is on lobby, but player data is on survival
then it will log out the player to lobby before server stop?
It just won't save where they disconnected from
That's the module which deals with reconnecting people to the server they left from
how to disable it? delete the the jar file?
Hello, does anyone know how to make the bungeecord kick message reflect spigot one? In spigot \n works but when I reconnect via the bungeecord network newlines are ignored
have to be weary due to how it works
But, basically, what's going on is it's mangling the components some how
(Really, you'd be better off dealing with bans on the proxy if you can too, tbqh)
(Really, you'd be better off dealing with bans on the proxy if you can too, tbqh)
That's what I was thinking, I think it's the only way to solve that
I don't know but seems bungee is stripping \n somehow in this function: https://github.com/SpigotMC/BungeeCord/blob/697f0875e65bc57676165b94654992f018e6851c/proxy/src/main/java/net/md_5/bungee/BungeeCord.java#L557
or maybe bungee uses another placeholder for newline?
use compoenents
let me try, thanks
It literally just calls toString on the component
It literally just calls toString on the component
bungee?
The issue is that all component detail is stripped, and clearly it doesn't reparse the \n's out
Yes
That's literally what that method you linked does
so no way to do that in spigot, i have to make a proxy plugin that handles bans received by the main spigot plugin?
Basically, yes
ok thanks, that's what i needed to be sure about
The issue is that there is no way to fix that stupidity without breaking other stupid things
the KickPlayer plugin message might accept new lines
but yeah, handling bans on the bungee directly is waaaay cleaner
eww, apparently I'm still using strings in my ban plugin
I'll add compatibility to each bungee ban plugin 🤦🏻♂️ oof, because for now i just execute /ban in the spigot server
it will be a pain to add compatibility to each bungee ban manager, but the only wae i think
what plugin is suggested as a bungeecord security plugin?
bungeeguard
thanks
or just use a firewall 
or firewall yes
Are there any recommended JVM arguments for Waterfall?
Check pins
Ahh, thanks
Waterfall provide anti exploit feature?
ReadTimeoutException means that a packet wasn't sent in so long
With mod packs, that's generally down to oddball issues with plugins being weird and bungee, there is an option in waterfall.yml that might help, metadata rewriting
yes
you can increase the timeout in config.yml, But, mods in general are a pain in the ass, so heavy level of "good luck"
how do i disable the built in /ip
i was using /ip to show the server ip and suddenly it stopped working and just gives "you dont have permission"
That's not a waterfall/paper command
yes
sweet thank you!
just want to take a second and say you all are great even with dumb little things like this. I appreciate it!
is there a way to stop people from joining a server directly without waterfall
all the parts run on one the same server its a linux server
Don't expose it to the world
bind on 127.0.0.1, or, the ptero interface IP if using ptero
nope i am not using ptero and if i bind it to localhost for some reason it still is accessible for outsiders
You are doing that in server.properties, and not just waterfall?
i have no access to the firewall the server is in a datacenter
if you have a dedi/vps, you have access to iptables
But, that's not 100% needed here, server.properties, set the ip to 127.0.0.1
If they can still see the backend servers, there is something wrong with your machines setup
localhost: or 127.0.0.1 is there a difference
no
ok
frik me then
i can still access the backend servers
and it runs on localhost:somthing
):
it does
config.yml
You literally highlighted the very setting for that
that's the timeout
that's 30 seconds
There are some things you can do, disabling entity metadata rewriting pulls us closer to velocity by a fair chunk, but, in general, bungee is somewhat bleh to work on
Some stuff in the pipeline just 10/10 health, but, a lot of stuff is at the wim of maintaining support for plugins in this fragile system :L
I am trying to start a brand new Waterfall server. Only changes I have made is the forced_hosts: Here is the config.yml https://pastebin.com/raw/wVJp1MTs I am now receiving a bunch of errors I do not know how to troubleshoot. Here is hat it says in the console https://pastebin.com/raw/qM91PZ8B
you have an error in your yaml syntax somewhere before the mentioned lines
yeah, you define both an empty section and section content on line 7 and 8
I mean, you have to remove child lines if you want to define an empty block ;)
I have done that and trying to restart the server now.
Well, no error but it does not look like a server started. Hmm.
hot question because i dont know anything. Plugins that use permissions have the permissions handled in the config.yml correct? im assuming so because even as an op on each server i cant use the commands.
you set the defaults of a bukkit plugin in the plugin.yml, yes. but this is #waterfall-help? 
okay
Does it matter i i appoint roo much ram to bungee?
uh, so the game version setting in the waterfall.yml doesn't affect which version can try to join the proxy? 
are there optimized bungeecord startup flags?
Hello, can anyone help me. One of my staff members is trying to join with 1.15.2 and gets this message. IDK if this is happening from the Paper Lobby Server or the Waterfall Proxy. EDIT: For some reason no one can join with 1.15.2 vanilla. Please ping me when you answered cause i am not constantly looking on discord.
@tidal musk Waterfall and BungeeCord channel rewrite logic is borked. If you need a quick fix, try Velocity, else perhaps I could cook up a Waterfall patch to disable this "feature"
Could you please do that? Cause i really dont want to change my servers proxy jar file due to plugins getting corrputed with other proxy softwares
Is there a way to allow bungeecord/waterfall to startup a server, or at least run a batch script on command?
I am using ExploitFixer plugin at every single paper server and HamsterAPI its dependency, i have bungeetablistplus plugin on waterfall, but rank doesnt show on Tab.
how to increase plugin messaging limit packet?
Hello, why there is such a long login delay with Waterfall Bungee?
It takes me 7seconds to login to the hub
whats weird is my waterfall (with the same config) that i use on my live server works fine, instant logins. but on my locally hosted testing bungeecord it takes like 5 seconds
and if it doesnt take 5 seconds it times out
i know - i was just reaffirming what the other person was saying
i came to post a diff question anyway
just saw someone else happened to be having the same issue and i thought id join in
tag me if yall respond. walking away and might forget to check back here
Hello, why there is such a long login delay with Waterfall Bungee? It takes me 7seconds to login to the hub ?
no waterfall does not have a long login delay
if its long, its anothers plugin fault
Well @wooden eagle we are not something like a paid helpdesk and not obliged in any way to give you any help at all or even a direct fix. In general we do like to help, but getting mad because we can't tell you a direct fix won't improve you chances.
No one got mad lol, I said I was looking for someone to help me if he knows the issue fix
test locally without any plugins to see if it's something with waterfall or your setup I guess
I hope you realise that linking to a single discord message instead of just copying/screenshotting it lowers the chance of someone taking a look...
xd
No plugins installed and I have the same issue as Shanny
If it means anything, the bungee is installed locally (not portforwarded) and linked up to one server atm
Its weird, because waterfall works perfectly on my live server
just on the local test server its wicked slow
I mean, do you run it on windows?
nope
or a potato xD
no sir
the same machine runs some other servers just fine
not a bungee tho
its just runs paper servers
for a test I suggest turning everything off that is not necessary to do the test
Is there a way to Disable Bungee commands/permissions
is the server external ip address
So, either you got some janky network setup, or, you have something opening a connection to your server on the machine
You'd need to work out what that is and get rid of it
Recently I've started thinking about splitting a survival world into 3 separate servers (overworld, nether, end).
I've tried to read up on that (haven't launched a test server yet), but I couldn't find anything useful.
Any tips on how to do it so that from player point of view it works like before? So that nether/end portals work normally etc.
I need just a nudge in the right direction.
@unkempt hinge never tried this, but when you create a multi-server setup you need to keep in mind that things like achievements and inventories need to be synced between the servers, so databases need to be setup etc etc
very plugin heavy
Yeah. I know there will be a bit of work to migrate everything to use MySQL and fill in all the gaps that vanilla normally handles.
I'm preparing myself for a fun ride :)
fun you say 😛
now as said i never did any of this, but you would probably need something like a proper bungeechat plugin
I'd go with mariaDB over mysql, but that is a personal preference ( same database, one is open source the other isn't )
probably vault on the minecraft servers themselves
mysql is free software, not open source 😛
I meant MySQL as in the type, I have more experience with Mariadb anyway (still have my Ansible to set it up somewhere).
how is it not open source? o.O
But I don't think I'll bother and will just have it hosted.
waterfall is just a fork 😢
I mean, it is
licensing
It's still got many of the stupid issues derived from being ontop of bungee
I think dorbian is right, from what I remember.