#waterfall-help
1 messages · Page 27 of 1
which one
permissions:
default:
- bungeecord.command.server
- bungeecord.command.list
- bungeecord.server.lobby
- bungeecord.server.pvp
- bungeecord.server.vd
admin:
- bungeecord.command.alert
- bungeecord.command.end
- bungeecord.command.ip
- bungeecord.command.reload
- bungeecord.server.developer
- bungeecord.server.prison
oh
i found it
Test it and make sure. lol
permissions:
default:
- bungeecord.command.server
- bungeecord.command.list
admin:
- bungeecord.command.alert
- bungeecord.command.end
- bungeecord.command.ip
- bungeecord.command.reload
This is what the default one looks like, I don't actually see an error there.
🤔
https://pomf.nachtraben.com/u/IQU1id2.png
Unless you had an error above the permissions what you sent parses correctly as a collection.
Most definitely.
Tried using a Bungee plugin on Waterfall and got an error, I was told that Waterfall should support Bungee plugins but I don't know if that's true. Here's the source of the plugin and the full Waterfall log: https://github.com/pmpy/SendHub
http://paste.ubuntu.com/p/bkQJMcHw7T/
not a waterfall issue
Am I correct on assuming that Bungee plugins should work with Waterfall though?
yes
Thank you.
But, that is a bug in the plugin, they're tryna interact with a null value
Thanks, I read the log and saw line 44 and saw that's where it's supposed to, I believe, insert the fallback server and message, I'll see if I can find someone to help me fix it or contact the dev!
someone was able to crash our bungee yesterday (https://www.youtube.com/watch?v=gB1HWkEeNEA) i opened up the logs and saw this InitialHandler - bad packet ID, are mods in use!? Did not read all bytes from packet class net.md_5.bungee.protocol.packet.Handshake 0 Protocol HANDSHAKE Direction TO_SERVER being spammed from a bunch of diff ips and our cpu usage was thru the roof. basically a ddos but is there anything i can do to mitigate it?
(DiscordBot) MUNCHYMC vs MCATTACK - length 2m 16s - 2 likes, 0 dislikes (100.0%) - 11 views - MC Attack on 2020.02.19
tcpshield is offering free protection
I got so burnt on ProxyPipe back in the enchanted days that I swore to never touch another anti DDOS service >~>
tcpshield is actually good
Basically, I do need to spend some time looking into toying with stuff, I just personalyl don't have the capacity to test stuff properly with the hardware I've got, so like.. rip
not all attack methods are public, and getting a setup enough to actually test stuff is a royal PITA that I kinda don't have much interest in doing
those attack only affect bungee/waterfall/proxy ?
to my understanding, yes
Saying that, these are generally attacks where the proxy is what's exposed to the world
There are much easier attack vectors for the server itself
I have the infrastructure and capacity to test, but we don't have these attacks at the network.
.g tcpshield
(DiscordBot) https://github.com/TCPShield -- TCPShield · GitHub: "TCPShield has one repository available. Follow their code on GitHub."
i use it. and alot of large mc network also use it
where's the catch
yea ikr
Same, this ought not be free..
The monetize your traffic and sell it big corp :^)
I use it, its legit
thats not the point
i belive they have said that they use it to learn about new attacks, dont quote me on that tough
Also they take donations
is there a way to forward mcserver motd to be display on bungee?
You are looking for ping_passthrough in config.yml
Note that it forwards the entire response to the client. If you really just want the description you're better off trying to configure the MOTD on the proxy.
"entire response to the client" wdym like version/icon/motd ?
yes
aight since my motd plugin doesnt send through bungee prob gonna use slp
Hello, new to the server. I joined with a problem I had.
Problem: Zero-Tick farms aren't working
Oh, now that I think about it, I only tried making it in a 1.15.1 world. Right now, it's 1.15.2
zero tick is configurable in paper.yml
ctrl-f "zero-tick"
we fix that bug by default but you can enable it if you want
waterfall tho lul
mods
Does 512MB is good enough for 128 players?
should be
Project Waterfall build #321: SUCCESS in 3 min 17 sec: https://papermc.io/ci/job/Waterfall/321/
Project Travertine build #130: SUCCESS in 1 min 29 sec: https://papermc.io/ci/job/Travertine/130/
Project Travertine build #131: SUCCESS in 1 min 28 sec: https://papermc.io/ci/job/Travertine/131/
Project Waterfall build #322: SUCCESS in 3 min 6 sec: https://papermc.io/ci/job/Waterfall/322/
this is what happens when you don't git diff before pushing 
no u didn't 
I saw that the rev had changed along with the version, just assumed it was typical git being git
Project Waterfall build #323: SUCCESS in 3 min 0 sec: https://papermc.io/ci/job/Waterfall/323/
Project Travertine build #132: SUCCESS in 1 min 24 sec: https://papermc.io/ci/job/Travertine/132/
Not sure if I should ask in here, but I've been having a weird connection issue. When some people connect they get through the load screen in the hub server but nothing loads and everything until they disconnect. I removed all plugins from both Waterfall and Paper and the issue still happens. The weirdest part is that there's no stacktraces when this happens. It simply gives the normal disconnect message for them on consoles and gives the player the "connection was closed by remote host" message on their client. The people with this issue don't have it on other servers and I am not sure how to approach this since it looks like it's an issue on Waterfall or Paper.
I've also seen electric's posts that it could possibly be a plugin handling packets badly but it happens when there's no plugins running too.
Hello, everyone in my server is invisible, i do not use plugins, the problem is not happening without waterfall
do you have plugins on waterfall?
no
i've removed it all to try
I don't have plugins in my paper or waterfall all empty
and it's logs from paper
honestly, no idea
This is weird af, it says disconnected, like server kicked the player but in fact i'm still in server and i can make changes, use commands etc
[23:14:05] [User Authenticator #1/INFO]: UUID of player ABeing is fcac4bb6-a8a0-38d4-b632-25c9d7649831
[23:14:05] [Server thread/INFO]: ABeing[/172.18.0.1:47252] logged in with entity id 11 at ([world]-1.3221416149453193, 79.0, -3.451840519356422)
[23:14:12] [User Authenticator #1/INFO]: UUID of player xBcee is 075660b3-6048-3934-be43-9b81c110c7d3
[23:14:12] [Server thread/INFO]: xBcee[/24.133.1.191:57746] logged in with entity id 12 at ([world]-3.2760705082073263, 79.0, -4.718133836693137)
[23:14:19] [Server thread/INFO]: xBcee lost connection: Disconnected
[23:14:19] [Server thread/INFO]: xBcee left the game
[23:14:54] [User Authenticator #1/INFO]: UUID of player xBcee is 075660b3-6048-3934-be43-9b81c110c7d3
[23:14:54] [Server thread/INFO]: xBcee[/172.18.0.1:47672] logged in with entity id 13 at ([world]-3.2760705082073263, 79.0, -4.718133836693137)
[23:15:14] [Server thread/INFO]: ABeing lost connection: Disconnected
[23:15:14] [Server thread/INFO]: ABeing left the game
[23:15:14] [Server thread/INFO]: xBcee lost connection: Disconnected
[23:15:14] [Server thread/INFO]: xBcee left the game
[23:15:43] [User Authenticator #1/INFO]: UUID of player ABeing is fcac4bb6-a8a0-38d4-b632-25c9d7649831
[23:15:43] [Server thread/INFO]: ABeing[/172.18.0.1:48020] logged in with entity id 14 at ([world]-5.332357906538114, 79.0, 2.666960819539068)
[23:15:50] [User Authenticator #1/INFO]: UUID of player xBcee is 075660b3-6048-3934-be43-9b81c110c7d3
[23:15:50] [Server thread/INFO]: xBcee[/172.18.0.1:48024] logged in with entity id 15 at ([world]-3.2760705082073263, 79.0, -4.718133836693137)
[23:16:00] [Server thread/INFO]: ABeing issued server command: /tp xBcee
[23:16:00] [Server thread/INFO]: [ABeing: Teleported ABeing to xBcee]
sounds like an old server bug
offline
People used to abuse the servers login handling code to force themselves to be invis
How can i fix this?
I don't recall if 1.12.2 was affected by that, probably was
either find a plugin or enable online mode
enable online mode on the proxy
afaik, that /should/ work, but, otherwise, you're on your own
if you're talking about enabling the online mod from waterfall configs
it's already enabled
then you've not enabled ip-forwarding properly
Does anyone know what I should do about my issue? Is there a way to check if it's a Waterfall or Paper issue?
start by describing your issue
our magic orbs aren't that powerful
xD I wrote it earlier
ah, i c
But basically when someone connects they go passed the loading screen but the world never finished loading up
I removed all the plugins on Waterfall and Paper to see if it fixes but it hasn't
i can only say that players have too high latency to your proxy very likely
or your host is having other issues (e.g some OVH dedis have networking issues)
So it might be the host after all
starting from bufferbloat and ending with other kinds of dark magic
Alright thank you, I just needed another opinion because I've exhausted everything, including updating Java, upgrading MC versions, swapping plugins, etc xD
Any known issue that can cause waterfall to have huge ping spikes?
poor gc configuration, leaky plugins etc.
Can i ask what's the difference between hexagon and travertine?
travertine is a fork of waterfall
Yeah, but from my understanding travertine and hexagon do the same thing
Not waterfall, hexagon
The difference is waterfall
hexagon bases their patches ontop of bungee, travertine bases ontop of waterfall
it's really a case of "do you want waterfalls improvements or not"
hexagon : travertine :: bungee : waterfall
mikro, ram usage is not over the top and neither is CPU usage
according to what?
Guys what does this mean??
[02:53:56 INFO]: [/192.168.1.1:54712] <-> InitialHandler has connected
[02:53:56 INFO]: [johncsuti|/192.168.1.1:54712] <-> ServerConnector [lobby] has connected
[02:53:57 INFO]: [johncsuti] disconnected with: Invalid payload REGISTER!
[02:53:57 INFO]: [/192.168.1.1:54712|johncsuti] <-> DownstreamBridge <-> [lobby] has disconnected
[02:53:57 INFO]: [/192.168.1.1:54712|johncsuti] -> UpstreamBridge has disconnected
i guess what im asking is is this done by waterfall or not?
Are you using a vanilla client? If not, try using vanilla.
using forge on vanilla i can join just fine
I need this to work so i can have vanilla servers with skyfactory 4 on the same network
It can be an issue with your lobby plugins (I've run across an issue in ViaVersion myself)
not a waterfall thing, that's gonna be the lobby server kicking you as craftbukkit has a limit on plugin channels
there is a flag to override that
What may causes Waterfall uses over 1000% cpu?
How to find the cause?
use a profiler or something, see what the thing is doing
hi
for the thing that says "could not connect to a default or fallback server" can i change it?
or how do i make a "fallback" server
i think that message is oudated since now the system used priorities
greetings.
i am running waterfall on windows and as you might know, when you have multiple servers each server takes an own command prompt window.
is there a workaround or a solution for that? It's kinda annoying.
^ dont think so.
How to fix the lag cause by this?
https://media.discordapp.net/attachments/399093265468489757/681436603935227904/unknown.png
er
follow up the trace and see where it's coming from?
Something tried to open a connection to something
what tool is that @tall stirrup
YourKit.
Coming from Netty IO thread.
is the initial waterfall install the same as with bungeecord?
or did they something change i have to consider?
same was bungee
thx
hey what is the most stable version to use with 1.8.8 clients ?
latest
the thing is that when i build my plugin and upload it to the proxy and restart it it uses the old one it's weeeeird af
must be some leftover plugin jar
e.g same plugin but with different file name
or you just copied it to the wrong location
mikroskeem thank you that was the problem after i renamed my plugin it's working the only thing now is that i have two versions in my server and only one in the FTP xp
fixed it it was named .jar so winscp couldn't find it
I am thinking to switch from velocity to waterfall
and what holds you back?
bungee = waterfall
yes
We have one or two things patched up
But, there still needs to be some work on it
as well as the whole fact that anything exposed to the internet is vulnerable to attacks
a crash is usually not a security issue
Basically, we have one or two things patched so it's slightly harder to overload the proxy, but, there is still work that needs to be done, just, that's work
thanks
here is any wiki
for config.yml?
The home of Spigot a high performance, no lag customized CraftBukkit Minecraft server API, and BungeeCord, the cloud server proxy.
Is there a way for bungee/waterfall when you restart a proxy you could have another one for fallback? So it doesn't kick all the players?
I was looking redisbungee but I couldn't find anything about it with having 2 servers if one goes down or anything, just communication between all servers
hmm
So you can have 2 proxies? 🤔
well it's not very simple to do but you can
You could probally just make the first proxy think that the 2nd one is a server, not too sure though.
if your waterfall network is using localhosts for each server with different ports....can you still use votifier
yes
yo legit....does nobody know about NuVotifier....cause I'm having trouble finding someone who knows something about this
read the documentation?
.g nuvotifier github wiki
(DiscordBot) https://github.com/NuVotifier/NuVotifier/wiki/Setup-Guide -- Setup Guide · NuVotifier/NuVotifier Wiki · GitHub: "A fork of Votifier, with more robust code and vote forwarding - NuVotifier/NuVotifier."
how do you know if you should Proxy based vote forward or Plugin messaging based forward
ok no no
so I'm for sure proxy based..
it supports plugin based automagically
it's not "automagically" for me XD
So...test votes through Minecraft server tools are working "successfully" but through planet minecrafts Votifier it's not
rsa key set up wrongly?
I don't believe so.....I could try the other servers keys...but it worked for mctools
yeah still nothing...
@rotund olive no idea if planet minecraft does anything different than usual sites but makes sure that the port and stuff is correct
@FlexstarSucks you'd have to run a separate server on that port while your proxy is down which is a bit tricky
it's the exact same as I'm putting into mctools
which is the part that's making no sense to me
prob change in
fallback_kick=\u00a7cCould not connect to a default or fallback server, please try again later: {0}
messgaes.properties
and recompile it
no you don't need to recompile/repackage jar
just extract the messages.properties into same dir where jar is
also @cold notch fyi ^
Sense the server was on a vm it had it's own ip address.....and instead of using the vms ip I was port forwarding my computers ip....works now :) thumbsup
Got it working
Secure waterfall
if you don't expose the backend servers to the internet it should be pretty secure
^^
since waterfall serves as the proxy server, you only need to port forward the port of the waterfall server, right?
yes
thanks
WHat version of waterfall do i need to use for a 1.7 server?
travertine
Yeah saw it just as i asked thank you
If I want to disable a specific module, can I just delete it?
orshould i remove it from the modules.yml
so both places
you need to delete jars as well yes
Hello,
I want to ask about if there's a way to grant permissions for players
you usually use a plugin for that like LuckPerms
No just i need to give permission like op
But when i use command.
/op
It's not found
@cold creek it's easy just register new glue record
And use the same ip
waterfall/bungeecord does not have the concept of op
there is a basic permissions system in the config but you really want to use a plugin for that
Ok ty
That answered nothing lol
Does anyone know how to make the server support both bungee connection and direct connection and bungee can forward the real IP🤔
"direct connection" not good !!!
ok i know but my server use authme login
if gonna run bungee all connection must go throught bungee unless u want the server to get fucked
oh ic offline server
yes
why make your life harder? just let everyone connect via bungee
that server did not use bungee before
and old client's server list contain the real ip of server
someone doesn't want to update their client
.g bungeecord config
(DiscordBot) https://www.spigotmc.org/wiki/bungeecord-configuration-guide/ -- BungeeCord Configuration Guide | SpigotMC - High...: "Apr 25, 2019 ... BungeeCord Configuration. Information for BungeeCord's config.yml ... motd: '&1Just another BungeeCord - Forced Host' address:..."
"someone doesn't want to update their client"
well 
what arguments do you guys use for waterfall?
@west arch What for arguments do you use for default BungeeCord?
i mean i just use java -Xms2G -Xmx2G -jar waterfall.jar
i just didnt know if there was more arguments
https://hastebin.com/eqotevates Welp this deaded my Network
The null ping attack method thats going around and cant be solved unless a better antibot?
Still needing to update ABD, but unsure if they solved this issue
Nah that wont solve it
Wait Protection for most "NullPing" attacks
Hmm seem like abd claim to protect from that
¯_(ツ)_/¯
Im seeing the read time out in proxy, but they are on one of my servers
Its like they are attempting to do something?
If it occurs again, will have to get around to updating ABD
Because it seemed from the old version to the new it had quite a few file changes
Wont have enough time in the next couple of days to do it sadly
at the end of the day, a piece of software can only deal with so much
Should Aikar flags be used for BungeeCord as well?
no
they can be, nut, just enabling G1 is generally fine
Hey, I know I'm asking this question without providing logs, but has anyone had issues where connection will randomly hand for a few seconds?
Seems to be at random, where all players will freeze for 1 - 2 seconds and continue normally
hi
does anyone know a 1.15 version of this?
very likely works with 1.15 as well
it doesn't
@tidal musk disconnects u within 2 seconds
i "found" a build for 1.15 it does'nt work either
it disconnects u immediately
any errors in console?
no
lemme try restarting the server and enable debug logging tho
this is the startup log
it doesn't log anything
Ask the dev for support they'll knownmore than here
that dev appears to be dead
but alright
ill try and get in contact with him
its open source
Is there a config
yes
its default
@viscid atlas basically what happens is that it works for like 5 seconds then i get this
but yeah ill get in contact with the dev
i found a fork
Hi I had my server setup then one of my stupid developers decided we are going to use Velocity and he ended up deleting my Bungee config. I've tried to remake it all again but I am having issues connecting. SSX can see my servers as online but if I type /server survival or use the Simple Portals I now get the following https://pastebin.com/QpuNEXsH
.g how to setup bungeecord
(DiscordBot) https://www.spigotmc.org/wiki/bungeecord-installation/ -- BungeeCord Installation | SpigotMC - High Performance...: "Jan 9, 2020 ... As a prerequisite, install Java 7 or above if you don't have it already (from Java.com). 2. Download the latest BungeeCord build from..."
The home of Spigot a high performance, no lag customized CraftBukkit Minecraft server API, and BungeeCord, the cloud server proxy.
read that
Yeah true good idea
u prob have ip or port wrong
well if u ´can conenct to lobby recheck other server in bungee.yml
okie
Still having issues so weird
so
apparently this doesn't work with the latest waterfall
Is it possible to run a waterfall proxy behind another waterfall proxy instance
When I tried that with bungee a while ago everything was broken
No
- The proxy prevents you from doing that
- Double rewriting packets is bleh, there is an option to disable entity metadata remapping, but, it's still kinda bleh
- I fail to see the gains of doing that
ah dammit, the gain is you could have multiple server networks on the same physical machine and if one needs a restart you can only restart that one
rolling restarts, etc, aren't new
Could someone help me understand the purpose of each setting in waterfall.yml
I havent found any documentation on that
.g bungeecord config docs
(DiscordBot) https://www.spigotmc.org/wiki/bungeecord-configuration-guide/ -- BungeeCord Configuration Guide | SpigotMC - High...: "Apr 25, 2019 ... BungeeCord Configuration. Information for BungeeCord's config.yml ... motd: '&1Just another BungeeCord - Forced Host' address:..."
The home of Spigot a high performance, no lag customized CraftBukkit Minecraft server API, and BungeeCord, the cloud server proxy.
Same config doc for waterfall
thankyou sir
We're trying to pull off a custom build for Waterfall, with success. However, when joining the 20w09a server we get the following error
Now I understand that this is all not supported. That being said, if anyone could point us in the right direction - much appreciated.
cc @eternal cairn maybe we can get the answer here 😛
tricky
@strong mantle this isn't a problem for waterfall to take, the proxy does what it should and works well, the mod doesn't as much
Hey can some dev look this up?
A bug when pressed tab in the console
Here is a photo:
It places two \
so the binding ip can it only be the 0.0.0.0?
What do you mean ? Do you want to test it ?
Not you just asking a Question
Ah
Hello, I am attempting to connect to a forge/sponge server from my spigot hub and it's not allowing me to connect to it.
I get the error message: "this server has mods that require fml/forge to be installed on the client."
I currently have forge_support set to true on the waterfall configuration and I am attempting to connect using the modded client that should work. I have all the mods necessary to connect.
you might need to connect directly to the forge server in some cases instead of going through a non-forge server before
kinda depends on the mods that are in use
There was supposedly a plugin that was used instead called SpongePls to circumvent this issue but that functionality was apparently put directly into Waterfall.
@strong mantle you're on 20w09a right?
And you opped yourself but then the op commands aren't working, that's the problem?
Try doing that, then relogging, and then seeing if they work.
It's happening on my server too so I think it's a vanilla thing, ops handling is messed up.
/op player returns an exception as well.
SpongePls was to fix an issue with FML and up forwarding together
It can't magically fix mods which require server side stuff, no idea how forge enforces that
No real changes between waterfall and hexacord that i'm aware of
So is it not possible to join a forge server from a vanilla server on waterfall?
I can connect to the server without bungee/waterfall with the same client
the required mods are there, it's just not recognizing it
god knows, would need looking into
It assumes I am on vanilla
are you connecting to the modded server first?
That's often an iffy one
The issue is is that you need to handshake with the client in order to get it to pass the mod list, that way we have the clients mods
I'd imagine if you connected to the modded server first, /server'd over to the lobby and jumped back, it would probably work
When I looked into it, it said Waterfall should be sending a FML reset packet
I didn't even know that fml had a reset packet
Honestly, most of the forge stuff is prior to my time
I need to get ontop of the latest forge stuff too, and honestly in my mind to look into rewriting the entirety of that mess
I mean, apparently Waterfall should already have this all built-in
Basically, bungee has forge support
md refuses to maintain said forge support to the point that it's no longer enabled there by default, bungees forge support is b0rked af, and we have odd tweaks here and there for it, but, there is a heavy understanding of stuff not being fond of the proxy in general
Part of the headaches now is that the code is half ass supported and unmaintained and really just needs a rewrite
Yeah I heard
This is what dualspiral said about the situation in 2017
Everywhere I try to find a solution, it's usually just saying to use Waterfall
and, we are tracking if we recieve that token
So, it /should/ be being sent to the client
I mean, the server
Hmm okay, so what does that mean exactly?
Yes
hm, do you have bungee mode configured properly in it?
On sponge?
yes
Yes
Basically, at this point, it's one of those needs testing type things
Either something with waterfall is busted there, or, something in your server itself ain't happy with the ip forwarding stuff potentially
Wdym by that ip forwarding stuff
The bungee is working fine for everything else
It's just the vanilla -> modded
that's causing the issue
I remember creating a sponge/paper network in the past, it is possible , you just have to configurate properly
(DiscordBot) All coords of large bases or bases with notable players at them have... : 2b2t - 8 comments, 29 points - Breithan, 5h ago
the exploit is fixed
on bungee or waterfal?
all you need to know is the exploit is fixed
🤔 mhh k
auth exploits in this day and age
or did someone just improperly configure something 
hypixel n 2b can understand 2b cuz hause high af but hypixel ¯_(ツ)_/¯
wat
Why there is 11 seconds between "InitialHandler has connected" and "ServerConnector [lobbyv2] has connected"
[14:06:51] [Netty Worker IO Thread #7/INFO]: [/xxx.134.172:15345] <-> InitialHandler has connected
[14:07:02] [Netty Worker IO Thread #7/INFO]: [anderson|/xxx.134.172:15345] <-> ServerConnector [lobbyv2] has connected
So I have to wait above 10 seconds with "Connecting to the server" screen.. Why so long?
idc what the image says, I know the exploit and I know it's fixed
Someone know whats going on? (Message above). Waiting above 10 seconds to connect to lobby server...
https://discordapp.com/channels/289587909051416579/555462214606389275/684019021066272768 how did it got fixed?
@tiny mica
yea I saw that but that is not enought to be sure it is fixed lmao
"or did someone just improperly configure something
" or outdated software
What exploit?
anyone could log into any account
I've run into an odd issue, my networks Waterfall instance can't seem to see a server on my network yet my computer can and so can a waterfall server running on my computer
recheck setting.
port/ip or only local ip enable or smt
are session servers down? https://pastebin.com/raw/0h9guQpK
.mojang
(DiscordBot) ****MJ|Session: **✔ **mojang.com: **✖ **MC|Session: **✔ **MJ|Account: **✔ **MJ|AuthServer: **✔ **MC|Textures: **✔ **api.mojang.com: **✔ **MC|Website: ✖
ah yes dark checkmarks on a dark background
i keep getting that for ppl
and using ping
is different IP
wow 2014
changing the java cache in the last few comments in that thread fixed it
any public plugin providing bungee request analytics (host used, amount of people pinging the server / joining the server, etc) you know of?
Guys, did someone use Spark on waterfall, so he could help me with proper use on that?
(please ping me on respond <3)
When using waterfall should I set the network-compression-threshold to -1 on servers that are running under the proxy?
I don't think so
You should always set it to -1 on servers behind the proxy.
If you don't you end up doing double the compression in some instances.
causing more bungee/server load instead of just bungee
Only time you shouldn't set to -1 is if your waterfall/bungee setup is on a network that can't handle the bandwith
HI, how are you
I'm currently using BungeeCord and i want to know if I should change it to waterfall.
If it ain't broke don't fix it
Waterfall has some patches to help with bot attacks, but that's one of the very few changes aside from code cleanuip
i thought it passed through? not doubled? i remember md_5 saying something like that
no it won't pass through
as bungee has to rewrite packets then it has to decompress server sent packets
one of the first elements in netty pipeline is compressor/decompressor, depending whether server requests compression or not
that proves the fact :p
And the amount of CPU usage it takes to do all of that work is insane. lol
when you have alot of clients connected
yeah
Almost every proxy (de)compresses packets
And it is very CPU-intensive
On the flip side, that compression can take place on all your cores since the Netty event loop will distribute connections over its thread pool
On the flip side, that compression can take place on all your cores since the Netty event loop will distribute connections over its thread pool
So proxies really do benefit from lots of cores and lots of bandwidth
Hi, is there a Dev about that Can Help Me please..
@leaden garden Is also working with me
.ask
If you have a question, please just ask it. Don't look for staff or topic experts. Don't ask to ask or ask if people are awake or available. Just ask the question to the channel straight out, and wait patiently for a reply.
does the paper server still read from the spigot.yml for bungeecord?
So to set the bungeecord: true in spigot.yml will work even if paper?
yes
ok. because paper is trusted for me
paper is based on spigot, that's why. every feature (and misfeature) what's present in spigot, is usually also present on paper unless noted otherwise
ok thanx! ill see how i go 🙂
Can some one offer a step by step to install a Waterfall based Hub server
As we are are having issues
yea, i just followed the guide and all i keep getting is no fallback server. its all on the same dedicated server so im just doing localhost
Any experts who have been working with BOT attacks? Recently someone has been sending bots to my server, and while my server won't let them in, these connections will lag out the proxy and will time-out people for a second. Is there anything that I could do or use against this?
no they can not connect
my player limit is 250 but the bots overload the proxy and it will timeout until the attack stops
and that has not happened before, I know the site that they used too
and its pointless to block ips because there's like 3000
Question, friend of mine is trying to convert a non-waterfall server to waterfall/paperMC and we can get in, but our inventories are empty and our UUID's aren't matching with our actual UUID's. Any ideas?
Yes, read the installation instructions, you didn't configure it properly
Somewhere in ur configs onlinemode isn't set right
It is helpful
ip forwarding
We followed the directions, set connection-throttle to -1, online to false on the servers and true on the waterfall server. What else?
Oh fml
I figured it out.
:v
hiii, just a question, i was wondering if anyone had any success with using waterfall with FTB? i can't seem to find anything on it >.< seeing as FTB uses 1.7.10 its rather annoying trying to add it to a hub lol
will that allow me to also use it on 1.15.x servers as well?
cat, you don’t have an idea how could I fix my issue ^^ ?
i mean i have an antibot plugin
but the connections lag out the proxy
It wont let them in to the server
How many connection/sec?
configure a connection limit on your firewall?
so while theres a bot attack others cant join?
at the end of the day, there is gonna be no magical silver bullet
load balancing is going to be your best overall option
that’d be how?
running several proxy instances to spread the attack vector
TCPShield is also free and offers mitigation
hi there, is there anyway to reload the config.yml without needing to restart the waterfall server?
greload
tyvm
Hello
I want to Ask about if there's a way to make my servers online mod and use waterfall proxy
no
spigot, etc, supports that using UUID/ip forwarding
There's a plugin?
No, it's bungee mode in spigot.yml
Read the installation instructions for bungeecord
he/she just kicked me
see #404345080917262369 for the rules

@tidal musk ok 
big f
Arrogant
if you continue to blatantly disrespect the rules you will be banned, last warning
any option to connect a forge server to waterfall without spongeforge? (not available for current version of forge)
1.13+ forge is unsupported
Bungee has dropped support for forge, and in general, I've just not gotten around to it/had the health/motivation for it
like 1.12.2 support is very outdated with forge
only goes to forge 2838 not 2847
You mean for spongeforge?
correct
I wouldn't be surprised if it work fine with 2847
Basically, what am saying is is that that is likely a non-issue
Have you actually tested it?
won't let me boot the server
Basically, I'm not aware of any reason it wouldn't work, that would often more signify an issue with another mod and sponge, vs sponge and forge itself; sponge is also generally fast on resolving those types of issues, etc
Might wanna provide logs/ask on their discord for some assistance
says it requires mixin 0.8 but its using 0.7.11
you have a mod using an older version of mixin
rename the sponge jar to something like asponge...
renaming the sponge jar with the prefix 'aaa' also doesn't work
best advice is to go ask on the sponge discord, also not really aware of any way that you'd get ip forwarding support otherwise
wasn't there a forge mod that added it or something along that line?
Project Waterfall build #324: FAILURE in 36 sec: https://papermc.io/ci/job/Waterfall/324/
Yippee, build fixed!
Project Waterfall build #325: FIXED in 3 min 9 sec: https://papermc.io/ci/job/Waterfall/325/
Project Travertine build #133: SUCCESS in 1 min 40 sec: https://papermc.io/ci/job/Travertine/133/
is this the "Technical" area for travertine as well?
sure
wait.. is that a person or a bot.. rather confusing lol
i guess not lol,
imagine discriminating against bots in 2020
okay here is my issue,
im running the latest Travertine on a 1.15.2 server, all other connections are fine, however, i have a 1.12.2 server running forge and SpongeForge, eveytime i try to connect to the 1.12.2 server i always get this error: the server you were previously on went down you have been connected to a fallback server then boots me, im 99% sure i have it all configured correctly, but i was wondering if anyone else had this issue or know a fix?
Forge Version: 1.12.1-14.23.5.2847
SpongeForge Version: 1.12.2-2838-7.1.8
are you connecting to the forge server directly instead of going through a lobby? (if not you might want to try that)
i have tried both,
when attempting direct connect it states i must enable IP forwarding in my Bungeecord config, which is set to true
i get the the server you were previously on went down you have been connected to a fallback server when trying through lobby
no, I mean directly connecting to it through travertine
e.g. by using forced hosts or a plugin to redirect to the server on login
how can i configure the forced hosts? and is there a plugin you could recommend?
there is settings in the config.yml for the forced hosts
and I have a plugin for connecting versions/forge to different servers, yes: https://github.com/Phoenix616/VersionConnector
i have tried the forced hosts but i still get the you must enable IP forwarding in your Bungeecord config
with the plugin, i have it installed on the travertine server is this correct?
|| [14:19:32 INFO]: [kyekye7964|/ip] <-> ServerConnector [the1122] has connected [14:19:32 ERROR]: [/ip|kyekye7964] <-> DownstreamBridge <-> [lobby] - encountered exception java.lang.IndexOutOfBoundsException: Tag type id 120 is out of bounds! [14:19:32 INFO]: [kyekye7964] disconnected with: The server you were previously on went down, you have been connected to a fallback server [14:19:32 INFO]: [/ip|kyekye7964] <-> DownstreamBridge <-> [lobby] has disconnected [14:19:32 WARN] [ViaBackwards]: Ignoring outgoing plugin message with channel: legacy:fml|hs [14:19:32 WARN] [ViaBackwards]: Ignoring outgoing plugin message with channel: legacy:fml|hs [14:19:32 INFO]: [/ip|kyekye7964] -> UpstreamBridge has disconnected [14:19:38 INFO]: [/ip] <-> InitialHandler has pinged [14:20:02 WARN]: [kyekye7964|/ip] <-> ServerConnector [the1122] - read timed out [14:20:02 INFO]: [kyekye7964|/ip] <-> ServerConnector [the1122] has disconnected||
Would this error be able to help at all?
When player count over about 130 Netty eat all cpu usage and causing lag.
YourKit report https://pastebin.com/ufcYhUUf
literally all your netty threads are working on opening socket connections?
Yes.
work out why
If those are opening connections, something is causing it to do that
No point tryna chase wild gooses, etc
Can you give me a direction to find out the problem?
work out what's opening those connections
netty might have some flag to offer more debugging on that
also, potentially likely unrelated, but worth a shot in the dark, disable the async dns setting in waterfall.yml if you haven't, in some oddball setups, that can ouch a lot
(god knows what causes that bar bad network setups or something, it literally uses googles dns servers last I knew o.O)
Is this causing by plugin or waterfall itself?
I've told you all I can
I got another report, is MasuiteWarp causing lag?
https://pastebin.com/ji8VHxG7
waterfall doesn't want to connect me to the forge sponge server
my magic 8 ball tells me that you need to provide more info
Yo, waterfall and bungeecord both "crash" (The player connection crashed) when I try to join my restriced server, 1 sec gonna upload log.
The target server is latest version of paper and this happens even when the target server is offline.
either something on your server mangled a packet, or there is a bug in bungees entity metadata rewriting
I can join the server directly, but not through bungee/waterfall?
You ruled nothing out there
Well nothing on the server could be faulty?
the client is known to ignore things like invalid entity metadata
Wouldn't regenerating the world help?
i fixed
How much ram does Waterfall need to be allocated?
500MB base and then 100MB per 100 players is a general rule of thumb
but plugins might change that
ok
like, if I have 11 plugins including ViaVersion, ViaBackwards, and ViaRewind, how much?
How do you rebuild the patches after updating upstream with ./waterfall merge?
waterfall rebuild?
oh
Wait
I already did that
but it didn't change the patches at all
I thought it's supposed to update the From TREE line in the patches?
then nothing significant changed
but for sure, here are the steps for upstream merge:
- ./waterfall m
- ./waterfall p & fix conflicts one by one if they appear
- ./waterfall rb
uh
not working
would these steps also apply to a fork of waterfall?
I made a script to make it easier
./hyperfall m
echo Merged upstream!
read -rsp $'Press enter to continue...\n'
echo Applying all the patches to Waterfall...
./hyperfall p
echo Applied all the patches to Waterfall!
echo Check to see if there is any conflicts and fix them if there are.
read -rsp $'Press enter to continue...\n'
echo Rebuilding the patches...
./hyperfall rb
echo Rebuilt the patches!
read -rsp $'Press enter to continue...\n'
using this:
why is it not working?
please just use a paste site...
I don't see what the conflict is...
it says in the message?
it literally says it failed at pom changes
even though there hasn't been pom changes for months
I based the script after what @tidal musk said:
but for sure, here are the steps for upstream merge:
- ./waterfall m
- ./waterfall p & fix conflicts one by one if they appear
- ./waterfall rb
What am I doing wrong?
yeah you have a conflict
you failed at 2nd step
see line 90
in paste
it literally says a conflict happened
something what git did not like
because the last time api/pom.xml was changed was last month
you better look into it manually
are the server names in config case insensitive?
yes
if you keep saying something in the lines of "impossible" then the confict will never get solved ;)
ok
I'll try
I need to modify the patches, right?
Cause when I try git rebase -i upstream/upstream as per Travertine CONTRIBUTING.md, I get
It looks like 'git am' is in progress. Cannot rebase.
you resolve the issue in the source, then use the continue command mentioned and at the end rebuild the patches from the fixed source/commits
oh
missed that
fatal: Resolve operation not in progress, we are not resuming.```
ugh...
I'm gonna start over with a fresh clone of my fork
Alright I did
I fixed the issue in the source and did git am --continue
and get:
fatal: Resolve operation not in progress, we are not resuming.
Install vault if you want economy: https://www.spigotmc.org/resources/vault.34315/
completely wrong channel
and also use #paper-help @tidal musk
now...
I fixed the issue in the source and did git am --continue
and get:
fatal: Resolve operation not in progress, we are not resuming.
why is that happening?
what did you do?
if git fails to apply single patch, then it stays in am mode
you fix the conflict
git add the-file
I did
Stage the changes?
yes, you need to stage the changes
also make sure that you are in the proxy repo, not patches repo
just did
and ran it again
and got the same result
fatal: Resolve operation not in progress, we are not resuming.
oh
also make sure that you are in the proxy repo, not patches repo
now it applied patches
🙂
you are operating with git repository after all
I was in the root directory
are you kidding me...
after all that
I ran my script again
and got the original issue:
Using index info to reconstruct a base tree...
M api/pom.xml
M bootstrap/pom.xml
M config/pom.xml
M module/pom.xml
M pom.xml
M proxy/pom.xml
Falling back to patching base and 3-way merge...
Auto-merging proxy/pom.xml
Auto-merging pom.xml
Auto-merging module/pom.xml
Auto-merging config/pom.xml
Auto-merging bootstrap/pom.xml
Auto-merging api/pom.xml
CONFLICT (content): Merge conflict in api/pom.xml
error: Failed to merge in the changes.
Patch failed at 0001 POM-Changes
Use 'git am --show-current-patch' to see the failed patch
When you have resolved this problem, run "git am --continue".
If you prefer to skip this patch, run "git am --skip" instead.
To restore the original branch and stop patching, run "git am --abort".
Something did not apply cleanly to HyperFall-Proxy.
Please review above details and finish the apply then
save the changes with rebuildPatches.sh
Applied all the patches to Waterfall!
Check to see if there is any conflicts and fix them if there are by fixing it in HyperFall-Proxy and running git am --continue.```
then I checked the source
and it had reverted back to having the original merge conflict
I made the script with the 3 steps you gave me. Is there a mistake in that?
Btw my fork is https://github.com/hypercubemc-github/HyperFall
Is there a way to fix the user IPs when using waterfall in pterodactyl?
then you must've made mistake somewhere else
running script again...
idk
did what i literally told you
applied cleanly
you must've skipped patch rebuilding
Alright got to the merge conflict...
I click use me on their changes
saved file
and then staged changes
and now if you run ./hyperfall p?
now go figure out where you fucked up
atom messed up 🤷♂️
smh my head indeed
Just a quick one.... if a Bungee /Waterfall server has an allocation of 30 slots, is that the max players on the whole network or the max connecting players to the Bungee server at one time? Thanks
is that the max players on the whole network or the max connecting players to the Bungee server at one time
what
whole network == bungee, no?
unless you have additional load balancing
your question is very ambiguous
Sorry.
Example: Bungee/Waterfall server set for 30 slots. 3 other servers after that set for 50 each.
Is the max players on the network determined by the Bungee/Waterfall server (30) or each of the servers that the players are connected to (potentially 150 - 3x50)?
the one with the lowest number
Excellent, exactly what I wanted to know. Many thanks! :o)
What does tab_complete: 4000 mean in Waterfall config?
if your proxy would allow 100 then a 50 slot server could still just hold 50
it limits the frequency in which tab completions can be send by the client and do stuff on the server in versions that have tab completion iirc
.g github waterfall tabcomplete patch
(phoenix616) https://github.com/PaperMC/Waterfall/blob/master/BungeeCord-Patches/0019-Add-basic-support-for-configurable-tab-complete-thro.patch -- Waterfall/0019-Add-basic-support-for-configurable-tab-comple...: "Subject: [PATCH] Add basic support for configurable tab-complete throttling. diff --git a/api/src/main/java/net/md_5/bungee/api/ProxyConfig.java ..."
doubt it
also what you mean aren't tab completions
they are command suggestions
waterfall might have a setting for that, but tbh: just fix your damn plugins, dude
oh okay, thanks
Does Waterfall block UUID Spoofing?
define uuid spoofing
waterfall DOES the Profile spoofing
that's absolutely needed for servers to get correct client ip address, name, textures etc.
Umm bypassing the proxy server with a cracked Minecraft account and force op with the Owner's "fake" account
.g lucko bungeeguard
(DiscordBot) https://github.com/lucko/BungeeGuard -- lucko/BungeeGuard: A simple plugin that verifies whether...: "BungeeGuard. BungeeGuard is a pair of plugins which intercept the BungeeCord handshake protocol, and allow backend servers to verify whether..."
there's your silver bullet
People say there's a way to bypass
And when I switch to paper, it breaks my server
Sending tons of dump
And crashing my server
there isn't a way to bypass unless you expose your servers to the internet
"bypass" well yes there always a bypass if user behind the keyboard is dumdum
boils down to that
yeah
Well, I use a website for a custom domain and it links to my forums and store
or if you don't use alternative measures to prevent external proxies from connecting to your backend servers...
Can someone send me waterfall tutorial in English
Is this a good place to ask about waterfall setup in pterodactyl?
.ask
If you have a question, please just ask it. Don't look for staff or topic experts. Don't ask to ask or ask if people are awake or available. Just ask the question to the channel straight out, and wait patiently for a reply.
refer to the bungeecord wiki for a standard install guide
as for ptero, bind your servers on to the pterodactyl0 interface
Ok, I have a problem where every player ip is 172.18.0.1 when the proxy server itself is run on pterodactyl, does anyone know how to fix that?
What's the difference between bungee cord and waterfall?
Servers themselves work fine but ip banning etc is kinda hard when everyone has the same ip
am guessing that you didn't setup ip forwarding/bungee mode properly
waterfall is bungee with extra patches on top, just like paper is to spigot
am guessing that you didn't setup ip forwarding/bungee mode properly
bungeecord = trueandonline-mode=false?
KKK thanks
yes, and on the proxy
the problem is that even the proxy gets the wrong ip
oh, god knows
if the proxy is getting the wrong IP, there is something weird with your setup
ok, I'll go poke the pterodactyl guys then Okay, I think I figured it out. my firewalld zone where the pterodactyl0 interface was connected had masquerade on. Disabled it and I got real ips on servers
pterodactyl <---- that the problem iirc there also another user thta got that problem today
centos+docker+pterodactyl is a multilayer mess where anything can go wrong, it seems
do u need a decent cpu for a proxy or is it pretty lightweight
depends on player counts, etc; But, it's pretty lightweight
scaling horizontally is recommended more over throwing a better CPU at it anyways
would something like a E5-1650 v2 work fine or would i7-7700K make a big difference
player count prolly won't go higher than like 20 tbh lol
the former would be much more than plenty then
kk thx
unless it's an oversold shared host and then cpu specs don't matter :^)
it is shared but idk about oversold
Hi, How do I get haproxy to work with waterfall while also showing correct client IP?
enable proxy_protocol in config.yml
what should I have in the haproxy config?
🤷♂️
any example for that? I currently have this but it isn't working listen minecraftsv01 bind 0.0.0.0:25565 mode tcp option tcplog balance roundrobin server mc01 BUNGEECORDHOSTIP:35061 send-proxy-v2
did I set something wrong or?
see how to configure haproxy
it's not that trivial but it's also nothing impossible
if haproxy is too hard then luckily nginx supports tcp proxying and haproxy proxy protocol as well
listen minecraft
bind :25565
mode tcp
option tcp-check
server proxy1 127.0.0.1:25566 check-send-proxy check send-proxy-v2
server proxy2 10.0.1.2:25566 check-send-proxy check send-proxy-v2 backup
@lone summit
I run waterfall and paper-clip, when someone join my server it says the server ip. i forgot how to make it so it seems to be from the right ip
So this is a thing for me. I tried a completely fresh waterfall config (just enabling ip forwarding and adding my servers to the list) and it still does it. And when it happens there is this error in the bungee console
uhhh, tricky scoreboard situation once again
Yea, is there a fix for it?
(@ me)
Hey I got this weird error that sent me back to my lobby. I'm using the latest version of waterfall and I wasn't doing anything in particular. Its not that big of an issue but I though I would report it. Error here: https://pastebin.com/pnNwfThY
well, are you using mods?
Where can I find waterfall maven repo?
dependency info is in the pom patch, I do need to get around to shoving that in the readme...
I take it I messed up somewhere?
try running a build/updating the maven indices
Ahaha. Thanks a lot.
https://www.spigotmc.org/wiki/spigot-plugin-yml-annotations/
Is there something similar for Bungee?
The home of Spigot a high performance, no lag customized CraftBukkit Minecraft server API, and BungeeCord, the cloud server proxy.
2 second google showed up that, no idea if it works or not 🤷♂️
it's just dumping out a plugin.yml, so literally shouldn't be an issue
Does Paper have a benchmark for players/CPU to see how many it can hold in a single server? I know there are many variables but just looking for an idea. Its an i7-4790K 16GB of ram
Thought i was in the Paper channel 
heyo so i just got a vps and got a server running
now I want to move the waterfall server onto it as well
would it be the same start.bat file that aikar recommends for bungee as it is for paper?
Use these flags exactly, only changing Xmx and Xms. These flags work and scale accordingly to any size of memory, even 500MB)
java -Xms512M -Xmx512M -XX:+UseG1GC -XX:+UnlockExperimentalVMOptions -XX:MaxGCPauseMillis=100 -XX:+DisableExplicitGC -XX:TargetSurvivorRatio=90 -XX:G1NewSizePercent=50 -XX:G1MaxNewSizePercent=80 -XX:G1MixedGCLiveThresholdPercent=35 -XX:+AlwaysPreTouch -XX:+ParallelRefProcEnabled -Dusing.aikars.flags=mcflags.emc.gs -jar waterfall-325.jar
Don't think it's really needed but yeah you can run it
so would just the regular start file be better?
java -Xms512M -Xmx512M -XX:+UseG1GC -jar waterfall-325.jar
I would recommend G1 at the least
Ill just use aikars flags and see how it goes then haha
Project Travertine build #134: SUCCESS in 3 min 5 sec: https://papermc.io/ci/job/Travertine/134/
Project Waterfall build #326: SUCCESS in 4 min 30 sec: https://papermc.io/ci/job/Waterfall/326/

Is it possible to have 2 Waterfall proxies for one network?
why not
do you have some sort of documentation that talks about that? I've only done single proxies before
you just have two seperate proxy servers with the exact same config, etc
literally just make two proxies connecting to the same backend...
DNS Round robin is generally used to "balance" them
help please 😦
ok sorry
smh
sorry guys .. 😦
So i cant quite figure it out but where should i be looking if players are only receiving plugin tab suggestion commands that are from the waterfall server and not the paper server they are on?
Depends on the version, but with mojangs new system, waterfall can't send its completions unless the server sends them; otherwise, on the older system, that's generally people disabling tab completion on the server itself, or some plugin breaking it
Ah ok
So am i correct if i say waterfall is basically forge with plugins?
no
Oh yeah am dumb
i was just wonderin cuz im about to make a forge server and i want plugins on it. the problem is Sponge is not even out for 1.14 is there some alternative?
theres things like mohist and all that, but expect issues, and no support
.g mohist
(DiscordBot) https://en.wikipedia.org/wiki/Mohism -- Mohism - Wikipedia: "Mohism or Moism was an ancient Chinese philosophy of logic, rational thought and science developed by the academic scholars who studied under the..."
1gb tops @mighty cloud
Bungee is fairly resource efficient since it doesn't hoard data and passes 95% of it directly down to the underlying server.
okey thanks!!
I've been looking places for resolution to the tag mismatch crash and I got lost
pls halp
er


