#waterfall-help

1 messages · Page 65 of 1

reef fulcrum
#

generally, bind the servers on the ptero interface and connect to the servers from waterfall using those IPs

tidal musk
#

its nothing wt the firewall

#

it is nothing wt that

idle nimbus
#

got strange problem some players which are from same city got big ping

#

and they time out

#

it happens randomly

#

no plugins on waterfall

#

and no plugins on paper side

#

1.17

#

server

#

it is hetzner dedicated machine

tidal musk
#

Sounds like it could be some sort of networking issue outside of your control, if it doesn't happen to other players

idle nimbus
#

all smooth

#

they are like 200km away of me so it could be there data centar

#

in there city

rigid nebula
#

Am i confused now or something, how do i set up the forced hosts?

#

Domain_name: targeted server?

#

It's basically not letting me join to the survival server using subdomains

rigid nebula
#

anyone?

reef fulcrum
#

hostname: server

#

only time it fails is if for some reason the client is sending a different name

#

(or if there is a plugin overriding it)

rigid nebula
#

Hmm

#

So subdomain.domain: servername

#

Idk what the client is sending, all i see logged is an ip adress

#

And I'm not sure if the reconnect plugin forces hub, i tried removing plugins too but didn't work

#

What about viaversion? Could that change the behavior?

royal hawk
# rigid nebula So subdomain.domain: servername

subdomain.domain.tld: servername
Keep in mind that:

  1. SRV records only work here if they point to an A-Record (and even then it might fail) and an A record itself is behind the SRV record
  2. If you use cloudflare you need to disable cloudflare for that subdomain
  3. Ports will be ignored
  4. CNAMEs will not correctly report, always use A-Records
#

For 4) Keep in mind that even flattened CNAMEs fall under that

rigid nebula
#

Hmm

#

Oh that makes sense

#

The A records are proxied through cloudflare

#

That's probably why then

royal hawk
#

That will break it

rigid nebula
#

I see

#

That's why then lmfao

#

Rip

#

I have an a record and a srv to my bungee

#

I guess ip lookup shows the vps ip then lol

royal hawk
#

Unless you pay for spectrum

rigid nebula
#

true

#

spectrum.. i guess it's more than 10 euro per month?

royal hawk
#

Well yes a lot more. Requires the pro plan at least + you pay for each gigabyte of bandwidth over 5

#

Not worth it

rigid nebula
#

ok so
A records points to IP
SRV record points to subdomain.domainname

royal hawk
#

Best scenario:
Your target is subdomain.example.tld
A subdomain > IP
SRV _subdomain > subdomain (A)

rigid nebula
#

oh

#

SRV is uh _minecraft

#

or wait

#

lemme copy it

#

SRV name is like the subdomain name
service _minecraft
protocol TCP etc

royal hawk
#

the issue is simple: minecraft doesn’t always fill the hostname field with the SRV record name so having the same name on the A record gives you that insurance

rigid nebula
#

the same name as in subdomain or the full domain name?

#

i did set the name on both the srv and a record to the subdomain's name

#

so do i have to change the service name from _minecraft to _subdomain?

royal hawk
rigid nebula
#

ah alright

rigid nebula
#

Thanks! It works

#

Btw any plugins recommended for waterfall?

tidal musk
#

how

#

acn

#

can

severe badger
#

By not using enter as delimiter

tidal musk
#

i make it so it show under my npc how many players it is in my otther bungee server?

upbeat wave
#

Any idea what might be causing players to get disconnected?
https://mclo.gs/94sBON4 - Is it AntiVPN at fault here or? because theres multiple errors.

reef fulcrum
#

I mean, come on...

#

You have a lookup for the session server, it times out

#

All your players then time out like, less than 10 seconds later

upbeat wave
#

Well i'm not that bright, i just wanted to make sure it actually was what i was seeing

#

thank you though

rigid nebula
#

What is waterfall changing what causes leaderboards to glitch out? Basically what's happening is that after a while no leaderboard updates anymore and they're fixed

violet hound
#

What does this error mean regarding my database?

#

And how would I fix

reef fulcrum
#

erm, you ran a bad query

#

or whatever caused that exception ran a bad query

violet hound
#

Well my issue that one of my plugins can't connect to my datbase

#

and I thought that might be the issue

reef fulcrum
#

probs bad config or something, idk

#

speak to the plugin author?

violet hound
#

it worked before, and then all of a sudden my database started acting up

pulsar python
#

hi

#

there is something like op in the bunge?

#

to have all the perms?

reef fulcrum
#

no, use a perm plugin

pulsar python
#

ok, and how?

pulsar python
reef fulcrum
#

don't mention

#

and what?

#

Install luckperms, etc, on your network, it lets you manage perms and even supports a wildcard, not that that is recommended, but, you can

pulsar python
reef fulcrum
#

Stop mentioning.

pulsar python
#

ah, sorry

reef fulcrum
#

The wildcard would be a wildcard for perms

pulsar python
#

is because i response

reef fulcrum
#

I recommend using LP for permissions, especially on networks, they have a wiki which goes over how to use the plugin and set it up, etc

tidal musk
reef fulcrum
#

what?

pulsar python
#

*i use it

#

luckyperms

#

but i have a issue

#

when i try to use commands i cant

#

essentials commands and a lot of plugins commands

#

but i does not recive any message

reef fulcrum
#

Which would have nothing to do with waterfall, unless you had some weird plugin on it

pulsar python
#

if i conect directly to the server i dont have that issue

reef fulcrum
#

if you can connect directly to the server, you've not configured it properly for one

pulsar python
#

but i set off the spigot bunge for tests

lost wharf
#

hi, my waterfall have issue when player join to the server they are drop down void new continue load terrain

royal hawk
# lost wharf

.paste your config please. What plugins are you running on the server and waterfall

marble kelpBOT
lost wharf
#

config for waterfall?

royal hawk
#

Yes

lost wharf
#

hold on

#

I using on waterfall
bungeeguard
multilobby
tab
and spark
luckperm

royal hawk
#

Okay, what’s the distance between waterfall and your server(s)

#

Like the physical distance

lost wharf
#

localhost

#

same machine

royal hawk
#

I have a feeling that this is the windows 10 loopback problem all over again

lost wharf
#

client problem?

royal hawk
#

Does it work with a vanilla client? Because I don’t think it would make a difference

lost wharf
#

it same thing

#

i try 2 client

#

and not different

#

it take 3-4s to load

#

and put me back to spawn

sacred furnace
#

Hi, im having some problems: i have setup waterfall on my pc and it was working all fine, but when i moved it to replit, it wasn't allowing me to connect to the server. What i have made wrong?

reef fulcrum
#

We don't support people abusing 3rd party services.

sacred furnace
#

"abusing"?

reef fulcrum
#

I highly doubt that the intent of replit is to allow people to run minecraft servers on.

#

Not to mention, such services often have weird setups and other quirks, so, you're on your own there

sacred furnace
#

im running the mc server in a mc server hoster

reef fulcrum
#

replit

sacred furnace
#

im using replit to host the proxy

reef fulcrum
#

just buy a cheap VPS

#

It's not the intent of those services to allow you to host stupid stuff on them, we do not provide support for what is basically misuse of these services especially when they often do stuff that prevents such things from happening

sacred furnace
#

ok

#

and how i restart a proxy from console?

#

it says cmd not found

reef fulcrum
#

there is no restart command

sacred furnace
#

oh

reef fulcrum
#

you need to stop it with the end command and then start it again

sacred furnace
#

ok, thx

jaunty grove
#

im having issues with my waterfall that i just set up

#

it connects me to the server just fine after some messing about

#

but in the minecraft client server list, sometimes it shows the bungee config set player limit of 1337, sometimes it shows the landing servers limit of 100

#

and if it does show the 1337 of the bungee, there is no MOTD. but if the 100 is there, it shows MOTD

#

just found out that if it shows the bungee player cap of 1337, the MOTD of the first entry of the "servers" list is shown

#

yeah, sometimes the MOTD and player count / limit of the target server "bleed through" 😐

marble kelpBOT
#

do you have ping-passthrough disabled?

#

and it looks like you disabled, odd

jaunty grove
#

its quite rare

#

maybe a bug

#

🙂

digital orbit
#

is the latest Waterfall build compatible with 1.8.8?

reef fulcrum
#

yes

digital orbit
#

ok thanks

slender bridge
#

is there any plugin can do console to discord, discord to console?
for bungee/waterfall

pulsar python
#

Hi

#

How many ram I need for 100 players?

marble kelpBOT
#

without plugins maybe 256MB or 512MB at most

tidal musk
#

Do all bungeecord plugins work with waterfall

shut cape
#

Should work

#

If it doesn't then you might reconsider using the plugin...

tropic wren
#

Send me download waterfall for minecraft 1.8 please

marble kelpBOT
#

.downloads

slender bridge
#

how to set admin?

slender bridge
young rampart
#

Install a permission plugin

slender bridge
young rampart
#

LuckPerms or BungeePerms

dull flume
#

Hey all, weird situation here. We updated most of our stuff back in July when MC updated. Since then I've been super busy with family issues and whatnot. Our server is getting some attention tho and I wanted to jump back in and update things and work out some bugs. I downloaded the latest Paper and Waterfall builds and threw them in. The proxy and backend server both load, but Waterfall cannot load PlaceholderAPI, Vault, or ProtocolLib... I've never seen this before. I backed down through about 5 versions of Waterfall to see if one of them would work and none do, including the version we had in before I updated today, and all give the same error. Looking through logs, it seems this issue only started 2 days ago on the 8th. Does anyone have any idea why these (up to date) plug ins suddenly can't be loaded by Waterfall? Thanks to anyone willing to assist!!!

nocturne fog
#

.paste the error you're getting please, that'll give a clue as to the reason

marble kelpBOT
prisma whale
#

Does PlaceholderAPI, Vault, etc even run on Bungee?

tidal musk
#

hello, anyone knows how I can add / in the messages.properties files without the slash being removed after restarting/loading the config

reef fulcrum
#

\/? maybe, forget how properties files work there

thorn ridge
#

hey all is there a guide or wiki on configuring haproxy with a minecraft server?

#

right now I'm getting this

#

oops how did I get in this chat

#

1 sec

dull flume
#

This is the old version, 443, that started throwing errors a couple of days ago. Won't load the 3 plugins I mentioned above. Of course I didn't know it was throwing errors until I tried to update today and started looking through logs.
https://paste.gg/p/anonymous/2cb317bf18b943a78a37cc4a7a0c400b

This is the first attempt to update Waterfall. 452 I believe, latest version.
https://paste.gg/p/anonymous/485a4c443c594bce8f3ea7de6ff6e12b

Thanks so much to anyone who can help. If I need to provide any more info please let me know.

nocturne fog
reef fulcrum
#

well, placeholder API is a bukkit plugin, not a proxy/bungee plugin for one

#

ah, yea, all of those plugins

dull flume
#

Well butter my biscuit and call me a carriage... I've had all three of these plugins on my proxy for over a year... I was sure I needed them for some of my bungee'ed plugins. I'll try disabling them and see what happens. Thanks for the input.

upbeat wave
#

I'm really not that bright in the arena of proxies so forgive me.
But how in the world can 127.0.0.1 get timed out?

#

players also got something like this

#

the 127.0.0.1 is continuing to get read timed out though :I

crimson fable
#

version mismatches, lag on server are the only two things i can think of that'd cause that

upbeat wave
#

the proxy is a lower version than the backend servers but that shouldn't matter?

reef fulcrum
#

read timed out means that the proxy didn't get anything, I kinda recall the InitialHandler being the clients side

upbeat wave
#

but when it happens to most of the players on the proxy? is it then client side still? D:

reef fulcrum
#

well, all it tells us is that it didn't recieve any packets in 30 seconds

#

that could be that the server isn't putting the client into a condition where it will send packets

upbeat wave
#

Do you have any troubleshooting steps you would go for?

reef fulcrum
#

plugins, diagnose network for packet loss, etc, inspect host machine for potential network issues (e.g. windows has a fun issue with the localhost bind setting thing sometimes)

upbeat wave
#

i think a call to hetzner is one of the thing i should priortize, because usually between 11 pm and 2 am theres ping spikes up to around 200ms

#

for all players

bitter jacinth
#

Hello ! I have a little using forge with Waterfall, when I connect to my server, the server disconnect me with this error :

zinc lily
#

does the server run forge >1.12.2 ?

bitter jacinth
#

not under nor up , it is 1.12.2

zinc lily
#

do you have the mods installed?

bitter jacinth
#

What a stupid question, yes I do.

zinc lily
#

any logs?

bitter jacinth
#

log from ?

zinc lily
#

waterfall, server

bitter jacinth
#
09:54:11 [INFO] Using zlib based native compressor.
09:54:11 [INFO] Enabled BungeeCord version git:BungeeCord-Bootstrap:1.17-R0.1-SNAPSHOT:6613aae:1607
09:54:11 [INFO] Not on Windows, attempting to use enhanced EpollEventLoop
09:54:11 [INFO] Epoll is working, utilising it!
09:54:11 [INFO] Discovered module: ModuleSpec(name=cmd_alert, file=modules/cmd_alert.jar, provider=JenkinsModuleSource())
09:54:11 [INFO] Discovered module: ModuleSpec(name=cmd_find, file=modules/cmd_find.jar, provider=JenkinsModuleSource())
09:54:11 [INFO] Discovered module: ModuleSpec(name=cmd_list, file=modules/cmd_list.jar, provider=JenkinsModuleSource())
09:54:11 [INFO] Discovered module: ModuleSpec(name=cmd_send, file=modules/cmd_send.jar, provider=JenkinsModuleSource())
09:54:11 [INFO] Discovered module: ModuleSpec(name=cmd_server, file=modules/cmd_server.jar, provider=JenkinsModuleSource())
09:54:11 [INFO] Discovered module: ModuleSpec(name=reconnect_yaml, file=modules/reconnect_yaml.jar, provider=JenkinsModuleSource())
09:54:11 [INFO] Loaded plugin reconnect_yaml version git:reconnect_yaml:1.17-R0.1-SNAPSHOT:6613aae:1607 by SpigotMC
09:54:11 [INFO] Loaded plugin cmd_find version git:cmd_find:1.17-R0.1-SNAPSHOT:6613aae:1607 by SpigotMC
09:54:11 [INFO] Loaded plugin cmd_server version git:cmd_server:1.17-R0.1-SNAPSHOT:6613aae:1607 by SpigotMC
09:54:11 [INFO] Loaded plugin cmd_alert version git:cmd_alert:1.17-R0.1-SNAPSHOT:6613aae:1607 by SpigotMC
09:54:11 [INFO] Loaded plugin cmd_send version git:cmd_send:1.17-R0.1-SNAPSHOT:6613aae:1607 by SpigotMC
09:54:11 [INFO] Loaded plugin cmd_list version git:cmd_list:1.17-R0.1-SNAPSHOT:6613aae:1607 by SpigotMC
09:54:11 [WARNING] Forced host server pvp is not defined
09:54:11 [WARNING] MinecraftForge support is currently unmaintained and may have unresolved issues. Please use at your own risk.
09:54:11 [INFO] Enabled plugin reconnect_yaml version git:reconnect_yaml:1.17-R0.1-SNAPSHOT:6613aae:1607 by SpigotMC
09:54:11 [INFO] Enabled plugin cmd_find version git:cmd_find:1.17-R0.1-SNAPSHOT:6613aae:1607 by SpigotMC
09:54:11 [INFO] Enabled plugin cmd_server version git:cmd_server:1.17-R0.1-SNAPSHOT:6613aae:1607 by SpigotMC
09:54:11 [INFO] Enabled plugin cmd_alert version git:cmd_alert:1.17-R0.1-SNAPSHOT:6613aae:1607 by SpigotMC
09:54:11 [INFO] Enabled plugin cmd_send version git:cmd_send:1.17-R0.1-SNAPSHOT:6613aae:1607 by SpigotMC
09:54:11 [INFO] Enabled plugin cmd_list version git:cmd_list:1.17-R0.1-SNAPSHOT:6613aae:1607 by SpigotMC
09:54:11 [INFO] Listening on /0.0.0.0:25575
container@pterodactyl~ Server marked as running...
09:54:16 [INFO] [/XX.XXX.X.XX:56164] <-> InitialHandler has pinged
09:54:20 [INFO] [/XX.XXX.X.XX:59179] <-> InitialHandler has connected
09:54:20 [INFO] [Eradium] <-> ServerConnector [lobby] has connected
09:54:20 [INFO] [Eradium] disconnected with: Kicked whilst connecting to lobby: This server has mods that require FML/Forge to be installed on the client. Contact your server admin for more details.
09:54:20 [INFO] [Eradium] -> UpstreamBridge has disconnected
09:54:20 [INFO] [Eradium] <-> ServerConnector [lobby] has disconnected```
#

Disconnecting Player: This server has mods that require FML/Forge to be installed on the client: [ArchitectureCraft, AutoRegLib, Better Animals Plus, Chameleon, Chisel, CoFH Core, CodeChicken Lib, CustomNPCs, EraMod, Extra Utilities 2, Extreme Reactors, Ido - 1.12 movement enhancement, Iron Chest, Magical Crops: Core, Placeable Items Mod, Redstone Flux, Secret Rooms 5, Storage Drawers, Sword Blocking Combat, Thermal Dynamics, Thermal Expansion, Thermal Foundation, U Team Core, Useful Backpacks, Varied Commodities, Zero CORE]

#

Every mod listed in the second message are all installed on server & client

zinc lily
#

interesting

bitter jacinth
#

Likely have found the error but It's a problem in some way

#

Okay, found out, but's It's really going to be a problem

#

The Ip forward stuff seems to break forge support

upbeat wave
#

Does connections between proxy and backend show up as "localhost"? or is this something else?

reef fulcrum
#

if connections show up as 127, then you're ever connecting from localhost or you've not setup IP forwarding

upbeat wave
#

i'm asking if the connections landing at the proxy, where it gets forwarded to backend (With correct setup and ip-forward)
Shows up as 127.0.0.1 or if its something else?

marble kelpBOT
#

pretty sure it will show up as whatever your proxy is bound to

reef fulcrum
#

I mean, if it's a connection over localhost then it's gonna show up on localhost?

#

We can't comment on your specific network setup

upbeat wave
#

i'm just trying to understand why the proxy would ping itself or whatever is pinging from 127.0.0.1

marble kelpBOT
#

what is even your exact issue?

upbeat wave
#

multiple things xD

#

random timeouts on players, random ping spikes

#

hetzner is ofcource denying all this so i'm just lost on what to do xD

marble kelpBOT
#

that sounds more like a network issues between the players and your machine, not the software

#

and I strongly doubt hetzner is having network issues, pretty sure they even provide you with health reports of their network somewhere, just look at that

reef fulcrum
#

I mean, that's what tools like mtr are for

upbeat wave
#

Well if its not something on the proxy nor the hetzner network - there shouldn't be any issues

#

theres nothing really in between other than tcpshield which are showing the "loss" is between hetzner and them

reef fulcrum
#

providers often suck at monitoring some aspects of their network, once again, mtr

upbeat wave
#

MTR blames hetzner

#

or atleast one of their nodes

reef fulcrum
#

have you showed it to them?

upbeat wave
#

yes many times

marble kelpBOT
#

so send that to them and move to a different node for the time being

upbeat wave
#

i can't move?

#

its a hetzner network node.

marble kelpBOT
#

tell them to

#

or move datacenters

#

or try to somehow route around it

upbeat wave
#

sounds expensive D:

reef fulcrum
#

I mean, if you can prove that the issue is on their network and they're/you're not gonna do anything about it, not much we can do, glhf

marble kelpBOT
#

yeah, just chargeback and use a different host lol

upbeat wave
#

i've had contact to hetzner multiple times, it always starts out as "We don't see any issue on our network - must be your ISP"

#

at one point i had a email trail on like 100 mails with them

#

where they null-routed me

marble kelpBOT
#

did you call them?

#

(ideally in native German :P)

upbeat wave
#

Yes, and he said "uhm we will forward this to the network team"

#

which works 9-18

marble kelpBOT
#

of course

upbeat wave
#

this was at 21

marble kelpBOT
#

I mean, you get what you pay for

upbeat wave
#

so that was waste of time

#

touché

reef fulcrum
#

I mean, you have been speaking to the networking team all along when asking about networking issues?

marble kelpBOT
#

you can't expect them to pay overtime with such low rates 👀

upbeat wave
reef fulcrum
#

depends on the day

upbeat wave
#

they don't even read the email trail, so i have to say the same things over and over xD

reef fulcrum
#

I mean, I ain't used their support system, but, most ticket systems let you chose what department to speak to when creating a ticket

upbeat wave
reef fulcrum
#

if you're speaking to the wrong department, it wouldn't be a surprise that they're half clueless and you're running in circles, especially given that things like gen support often acts as gate keepers

#

And onc eagain, was that the networking department?

upbeat wave
#

Yes it was

#

can't be a bottleneck if its all users that are affected xd

reef fulcrum
#

I mean, I can't really say anything based on that given I can't see the report

#

I mean, all it takes is for some 3rd party peer to have issues and shit goes sideways

marble kelpBOT
#

just send them info which clearly shows that it's not in your ISP

upbeat wave
#

I know, but this is a issue where everyone gets mostly the exact same latency

#

(200ms)

marble kelpBOT
#

and ideally tests from multiple different locations and servers e.g. ovh

reef fulcrum
#

At the end of the day, if the issue is on their side, you're basically gonna have to prove it, and sometimes that cna be a royal headache

#

all it takes is for a common route to be congested or some shit

upbeat wave
#

usually its their core24.fsn1 that is having issues

reef fulcrum
#

I mean, we're here on the assumption that you've read the mtr report properly

#

given that we can't see the report, we can't really comment on it, if it's an issue with their network, er, your options are either convince them that it is or go somewhere better, welcome to the joys of budget bin hosts

upbeat wave
#

i don't have a report on hand currently, but i'll make sure to save one when it happens again

upbeat wave
#

could pay for the server in like a couple of months

reef fulcrum
#

relative to the realm you're in? Yes, that's budget bin hosting

#

hetzner is a company who do it for cheap

upbeat wave
#

fair point

reef fulcrum
#

and are on the same price point as OVH more or less, OVH is literally the defininition of a budget host in this industry

upbeat wave
#

sorry forgot to remove the ping >.<

reef fulcrum
#

what?

upbeat wave
#

A Ryzen 9 5950x is "Consumer/Prosumer" grade hardware

#

not server grade

reef fulcrum
#

I mean, the rise of consumer hardware really says it all, these are not hosting providers doing the typicals of hosting, they're priced at a point for me and you, not "for the enterprise"

#

That means that you lose on on the 24/7 support you'd expect for an enterprise business, the # and quality of tech hands, etc, etc

#

am sure that they're a competent company but if you want "hands down your pants" support when shit goes sideways, your budget is too cheap

upbeat wave
#

the theory i'm going for currently is that their Stateless firewall is doing some wierd stuff and that might be the issue

#

i'm just shooting in every direction till something "sticks" i guess

reef fulcrum
#

Hands down your pants support implies that they'd be going super crazy to diagnose where the issue is and work with their peers if it's a peering issue

upbeat wave
#

saying its my fault on first message is not doing anything - they could have a bot do that

#

Could SQLite databases be the ones pinging the server?

marble kelpBOT
#

SQLite doesn't do anything on its own

upbeat wave
#

i'm so confused what the pings are then :-:

#

half of which are timing out

marble kelpBOT
#

some plugins do stupid shit like ping the server to obtain the motd or something

hazy sun
#

Hi anyone know some method to auto-restart waterfall instance every day or one time a week? Thanks

prisma whale
#

if windows, task scheduler + nssm or firedaemon

acoustic oriole
#

Hi

#

My waterfall server kicks randomly players with message read timed out or Connection reset by peer

#

Log

[22:13:40 WARN]: [/XXX.XXX.XXX.XXX:38084|PFilip] -> UpstreamBridge - read timed out
[22:13:40 INFO]: [PFilip] disconnected with: ReadTimeoutException : null
[22:13:40 INFO]: [/XXX.XXX.XXX.XXX:38084|PFilip] -> UpstreamBridge has disconnected
pearl blaze
#

Hmm

reef fulcrum
#

read timed out just tells us that the proxy didn't get any traffic in 30 seconds

acoustic oriole
#

Ok

#

but it happens randomly

#

like when player is on the server for 10-30 min

#

and it doesn't happen on normal bungee

pearl blaze
#

Hmm

royal hawk
acoustic oriole
#

:<

#

Ok

#

I just installed waterfall without plugins and the i got same error

royal hawk
#

This is on connection client <> proxy

acoustic oriole
#

so i think i will open bug on gh

acoustic oriole
reef fulcrum
#

Please don't

pearl blaze
#

🧐

reef fulcrum
#

There is nothing there that we can work off

acoustic oriole
#

So this is problem on my VPS/internet provider?

reef fulcrum
#

There is some weird issue with bungee in general that for some reason some people basically eat dirt with the proxy

#

One might suggest looking at velocity

acoustic oriole
#

Ok, I will check that

royal hawk
#

If you’re running this proxy in Linux chances are the modifications to the natives in waterfall are causing it, but that’s insanely hard to prove

acoustic oriole
#

Yes, i run it on Linux.

royal hawk
#

And that’s sadly hardly fixable. Most cheap VPS Hosters use hypervisors that cause these issues

acoustic oriole
#

The VPS provider is OVH

#

VPS wasn't cheap

royal hawk
#

Then I don’t see that being an issue

acoustic oriole
#

I will check velocity

royal hawk
#

Maybe something with the DoS protection OVH uses

acoustic oriole
#

Maybe

#

This problem started when i changed VPS location from FR to PL

#

I'm from Poland and ping in FR was very high

reef fulcrum
#

"very high"

acoustic oriole
#

Like 62ms

#

in PL i have around 16ms

reef fulcrum
#

I mean, that sounds around where I'd expect it going so far

acoustic oriole
#

Oh, ok.

acoustic oriole
#

So i checked velocity

#

and no error

royal hawk
#

Then stay with velocity

acoustic oriole
#

Also I checked waterfall logs and this read timed out error appeared couple weeks after VPS change.

#

So I think this isn't caused by VPS/firewall config

#

maybe ISP or Waterfall api or something

pearl blaze
#

Hello, I have a waterfall for some time everything was ok but recently for about 3 days it still crashes every 20m players that Read time out I bought eyfencord and there is also, as far as I know, it's based on waterfall so the fault is probably on your side, I would ask for repair, if there is one possibility

tough vale
#

Send logs

#

.paste

marble kelpBOT
hot wave
#

When I put / in my chat I get the commands of the Waterfall plugins even if I have in my mode that does not tab any command, is there a way to avoid that?

dawn umbra
#

why would i get expetion null when joining a modded server?

lost river
#

it is a ReadTimeoutException, not just null

dawn umbra
#

why would i get that

#

tho

#

it only happens when i try to join my modded server

lost river
#

You get connected and directly disconnected on the Revelation Server

#

do you have the modded server in offline mode?

dawn umbra
#

yes

#

it has ip forwarding

#

and it is bungeemode true

#

should it be online mode?

lost river
#

it should be offline mode if its behind a bungeecord proxy

#

however how did you enabel ip forwarding/bungeecord=true on the modded server?

dawn umbra
#

i use sponge forge

#

on modded server side

#

they have a config file i have to enable

#

those two settings

lost river
#

did you restart the modded server after changing that config file?

dawn umbra
#

ya

#

day before yesterday my players had issues joining modded server

#

so i struggled to get to join the server as well

lost river
#

is there an error message in the log of the forge server?

dawn umbra
#

when i removed it from proxy it allowed me to join the server

#

but when i added it to the proxy again i got this expetion

#

does not say i join that server even

lost river
#

did you make sure to input the right ip address and port in the proxy config?

dawn umbra
#

yep

#

it is on same machine

#

as proxy/bungee

lost river
#

so in the proxy config it says something like 127.0.0.1:port-of-modded-server?

#

do you have the server ip set in the server.properties file of the modded server?

dawn umbra
#

ya i had it that and i changed it to the ip of the machine

#

i thought it might be the ip but it was not it

#

but i will change it back

lost river
#

change both places to 127.0.0.1 as ip

dawn umbra
#

in the modded server i have it as 0.0.0.0

lost river
#

that should also work

dawn umbra
#

ok doing a restart

#

that was on modded side

royal hawk
# dawn umbra ok doing a restart

are you running the latest waterfall build? The plugin-channel mechanics changed with it, if youre on the latest build try the one before that and vice versa

#

this is from what I can see an issue related to plugin-messaging

dawn umbra
#

IDK i will download and upload it

#

cause atm i have my modded server disconnected on the proxy

#

so i can figure out if i can fix it

#

so my players can play

#

will do some testing later

#

i am having some headache cause of this problem

dawn umbra
#

now i got this

#

😭

dawn umbra
#

and got this error trying to join modded server

#

that is just to my hub

#

that it gave me that

#

also proxy just changed all this to that ip

dawn umbra
#

it is a if the modded server is not picked up on the proxy side

pearl blaze
dawn umbra
#

no i haven't figure it out

pearl blaze
#

It has been happening with me for several days and with you

dawn umbra
#

what i did was i made my modded server temporary independed meaning it is standing on it own and i gave it the port on my dns

#

i only got it today

#

ish

#

usually i got it 3/4 times on my side but for my players none

pearl blaze
#

For me it looks like that every now and then it crashes all players with this reason, it happens every 30m

#

I lost a lot of players because of this bug, so I'm trying to fix it but nothing works, I really tried everything

dawn umbra
#

i think it is a bug with the waterfall

#

i think we should open an issue on github

#

if it is not already created

acoustic oriole
#

Hi, I reinstalled everything on my VPS and I sill get error.
Waterfall:

[15:42:55 WARN]: [/XXX.XXX.XXX.XXX:45798|PFilip] -> UpstreamBridge - NativeIoException: readAddress(..) failed: Connection reset by peer
[15:42:55 INFO]: [PFilip] disconnected with: NativeIoException : readAddress(..) failed: Connection reset by peer
[15:42:55 INFO]: [/XXX.XXX.XXX.XXX:45798|PFilip] <-> DownstreamBridge <-> [surv1] has disconnected
[15:42:55 INFO]: [/XXX.XXX.XXX.XXX:45798|PFilip] -> UpstreamBridge has disconnected
pearl blaze
marble kelpBOT
#

@dawn umbra only forge up to 1.12 can be properly supported

dawn umbra
#

that is what i am using

pearl blaze
#

I think it's not because he is not the first person, and probably even 5th person, so the problem seems to me that he is on the waterfall side

dawn umbra
#

FTB revelation is the modpack i have

marble kelpBOT
#

also try disabling entity metadata rewriting, some mod stuff conflicts with that

#

(or directly use Velocity I guess)

dawn umbra
#

i think i have that disabled allready

#

or enabled

#

IDK anymore

#

it is in true

#

meaning it is disabled

pearl blaze
#

I don't know

dawn umbra
#

and if i do have the server linked to my waterfall and i try to join from hub it does not kinda pickup that that server is there but does pick it up

#

it does not send me to that server at all

marble kelpBOT
#

try using forced hosts or a plugin to connect to the modded server directly on join. some mods really don't like server switches

pearl blaze
#

Hmm, I don't want to mislead anyone but I don't think this is the reason why a few people have the same thing I did with kicking players, including me

dawn umbra
royal hawk
hot wave
#

ah :l

maiden sparrow
#

I restarted my dedicated server and now I can't connect to any of the servers

reef fulcrum
#

I mean, the proxy thinks that you're apparently already connected to the server you're tryna be sent to

#

if that's "it failed to connect and is sending me back to where I was and shouting about it", check your logs, etc

royal hawk
reef fulcrum
#

don't mention

#

no logs = no help

royal hawk
#

I want to say congratulations you found a bungee design flaw. I’m nearly positive it is the thing I’m talking about

#

This happens reliably when a connection is lost and the user switches but a plugin keeps making connection attempts one after another. If you time it right it will break the flow there and the server info will become stuck

maiden sparrow
#

And how can I fix it?

#

I even rebooted the dedicated server

royal hawk
#

Given this is the thing I am talking about it should be fixed by reconnecting to the bungee

maiden sparrow
#

even if I send myself to a server that it's offline I still get this

reef fulcrum
#

did you just log out and log back in?

#

can you reproduce without plugins on the proxy?

maiden sparrow
#

I have a plugin "Hopecommander"

#

I think this is the problem

#

nope

maiden sparrow
royal hawk
#

Is that a “cracked” auth plugin? We don’t setups with those or piracy

maiden sparrow
#

no

#

it's a login plugin

royal hawk
#

And the purpose of that is?

maiden sparrow
#

for login

royal hawk
#

Login is handled by mojangs session servers if you use legitimate versions of the game. Anything else is most likely piracy

maiden sparrow
#

do you know a way to fix this?

quartz mauve
#

Don't use weird auth plugins but the legitimate mojangs auth servers

dire plank
#

Is there anyway we can have a proxy -> proxy?
Not sure if I should post this on #velocity-help or #waterfall-help. Right now im using velocity, but I could change to whatever supports this:

How we want it:

#

Why: We have two servers merging together. Want specific proxy plugins to specific server.

#

Yellow: Main proxy
Red: Agonia SMPs proxy
Dark red: xFuns proxy

#

Please tag me @dire plank if you have an answer <3

eternal cairn
#

Neither Velocity nor Waterfall supports this.

#

Take a look at HAproxy if you don’t want plugins on the yellow one, though

#

it would potentially require a 3rd instance of Velocity if you need it to connect to the servers without going thru the red and brown ones

dire plank
#

No plugins needed in the yellow one

modest elm
#

Hi

#

i'm getting this warning when i start my server

[15:20:58 WARN]: Could not bind to host /0.0.0.0:25632
io.netty.channel.unix.Errors$NativeIoException: bind(..) failed: Address already in use
#

and then i get this when i try to join

reef fulcrum
#

that port is already in use by something apparently

modest elm
tulip surgeBOT
lost river
# modest elm this is the entire log
[15:20:58 WARN]: Could not bind to host /0.0.0.0:25632
io.netty.channel.unix.Errors$NativeIoException: bind(..) failed: Address already in use

There is already some program using that port.
If you have console access you can use netstat -t -l -p -n or something similar to check which program listens on that port.

royal hawk
viscid idol
#

Why are priorities & force_default_server broken? I have my priorities list as such: spawn, vanilla and force_default_server: false I want a player on first join to connect to 'spawn' and then have them log back in on whatever server they logged out on, but bungeecord is sending players back to 'spawn' always

#

I haven't used BungeeCord/Waterfall in quite a bit, has something changed that I'm unaware of...?

reef fulcrum
#

you didn't setup a forced host did you?

viscid idol
#

Nope

#

forced_hosts is an empty map

reef fulcrum
#

otherwise, disabling it should move you back to the server you where on when you left

viscid idol
#
  ping_passthrough: false
  priorities:
  - spawn
  - vanilla
  bind_local_address: true
  host: 0.0.0.0:25565
  max_players: 50
  tab_size: 60
  force_default_server: false```
#

It's just not remembering my logout server. I end up on 'spawn' every time... I tried disabling plugins on the proxy but that didn't change anything

#

It's physically saving it in the locations.yml, though. I have no clue as to why its not working

reef fulcrum
#

check the logs

#

make sure that the reconnect module I think it was loaded

#

best I got is check plugins

viscid idol
#

Yeah its loading and enabling it. I tried a run without proxy plugins and it's still an issue

viscid idol
#

Could it be because of routing through TCPShield?

#

I was infact TCPShield.

tidal musk
#

what is this (red is ip address of a player)

zinc lily
#

means a connection took too long to establish/respond

small fractal
#

why do people freak about their ip

#

i know it leaks a 100 km radius but why you freaking out

green sequoia
#

I mean it isn’t even theirs, it’s just their friends

#

Your own would be counted as lan ip

dawn umbra
#

ya now you have a network and you had some issue and you have about 50 different players from around the world connected and their ip show in console it is up to u to protect them rather than to have them be unsafe cause they can easily turn around and sue you for damages and it all also depend on where the server is hosted too

#

some companies has a strict policy towards safety

reef fulcrum
#

I mean, cutting out the IPs just makes it harder on us because now all we know is "read timed out"

#

I mean, I've already explained what read timed out means 20 times in the past month, but, like, for all we know those IPs trace back to some random host or some shit running on their own machine

dawn umbra
#

lol

marble kelpBOT
#

wasn't there a config option to set the max amount or do I misremember that?

reef fulcrum
#

No

#

I mean

#

There was literally already an open issue report for that.

dawn umbra
#

welp this is casuing me to have my modded server dismantled from proxy server. kinda annoying

reef fulcrum
#

oh, wait, maybe not exactly a duplicate, just, some common major issue there was already fixed

dawn umbra
#

give me the fixed version then pls

reef fulcrum
#

don't mention

dawn umbra
#

cause atm latest version is F'ed

reef fulcrum
#

Please actually read what I said

dawn umbra
#

you said it was not exactally the same but there was a fix

#

so i am asking for the fix

#

cause i do not see the fix

reef fulcrum
#

oh, wait

#

some common major issue there that was fixed

#

maybe not exactly a duplicate

dawn umbra
#

well then why did you close my issue

#

if it is not fixed yet

reef fulcrum
#

Because I thought it was a dupe of the existing issue given that 99.999% of them are

dawn umbra
#

well i would recommend opening the issue again so that issue can be resolved please

marble kelpBOT
#

bungee and I think the Minecraft client itself can only register 128 channels

dawn umbra
#

cause i am not going to open new one

reef fulcrum
#

I was literally just working on a thing for it now

#

but, i changed my mind, i cba

dawn umbra
#

cause velocity does not meet my touch atm there is not plugins i can use to make the exact same things i am using now

royal hawk
#

A patch to add the same -Dpaper.disableChannelLimit arg paper has would fix it

reef fulcrum
#

Bearing in mind that I proposed the fix for the original issue like literally tiiime ago

marble kelpBOT
#

oh I think I thought of the paper option lol

reef fulcrum
#

just wanted a config option to set the max or -1 to disable

#

let people shoot themselves in the foot if they want or work out a sane limit for themselves

royal hawk
#

Velocity default is 1024

#

fine throw another config option in the waterfall.yml

reef fulcrum
#

I mean, I'm iffy on sys-props

#

Because, I love putting the "feel free to shoot yourself in the foot" options in system properties, but, then somebody shoots themselves in the foot anyways and cries, or somebody has a valid usecase for it and is using a shared host so it's all fun

dawn umbra
#

i learn by shooting my self in the foot

#

so i break first then build afterwards

indigo dune
#

with newest waterfall i get 09:07:45 [ERROR]: java.lang.NullPointerException: Cannot invoke "java.lang.reflect.Method.invoke(Object, Object[])" because "com.viaversion.viaversion.bungee.handlers.BungeeServerHandler.getRelayMessages" is null now, maybe its the same topic? dont know if i have to talk with via about this...

reef fulcrum
#

that sounds like via needs updating

#

but, I mean, can't really say, there is not enough info there, but 🤷‍♂️

royal hawk
#

Hurrr durr @frank coral fix

frank coral
indigo dune
#

nice thanks

royal hawk
royal hawk
#

If it does ignore it

frank coral
#

you're mistaken then

royal hawk
frank coral
#

wat

royal hawk
#

Ah I see how you did it

#

That’s stupid but alright

indigo dune
#

🙂

keen vault
#

I'm currently trying to put a SevTech-Ages as well as a GT:NH-Server behind Waterfall.
When trying to connect (tested with the ST-A-Server) I keep getting the error in the attached image.
forge_support is set to true in the config.yml and I can connect to the servers without a problem when not using Waterfall.
Is there something else I need to configure to make this setup work?

keen vault
quartz mauve
#

Only forge up to 1.12 is supported

keen vault
#

So there is no way to make it work?

royal hawk
#

That is a design “flaw” in bungeecord

royal hawk
keen vault
#

I'll try it and give you feedback as soon as I'm done

#

The option appeared and I now have set it to 1024 - currently starting ST-A (could take a few minutes as I'm on my laptop)

#

Works like a charm!
Vielen lieben Dank 😄

hot wave
#

Where do I download the bungeecord messages? I think they have implemented a new one that I don't have translated.

#

I don't remember where I downloaded them

severe badger
hot wave
#

This message does not come.

#

"The server is currently under repair, try later..."

reef fulcrum
#

that's not a waterfall message

#

That's the message which is being sent from the server

#

i.e. a kick message from the server is in english, but, the string which that message is shown in on the proxy itself, you're translated already

hot wave
#

oh ok, sorry

#

thx

keen vault
reef fulcrum
#

Waterfall hasn't supported 1.7.10 in years

#

so, generally, glhf there

#

may wanna look at velocity or maybe fork travertine and maintain it yourself 🤷‍♂️

keen vault
#

Yeah… I'll look into it

dawn umbra
#

Thx

dawn umbra
#

Hello
the message channel thing that was added, after changing the values it still gives me same messages over again. just to let ya know about this

#

when i had it changed i got the readtimeoutexpetion: null again to the server but other player got the >.... [19:52:12 INFO]: [MrWhiteFlamesYT] disconnected with: IllegalArgumentException : Channel name too long. This limit can be configured in the waterfall.yml @ com.google.common.base.Preconditions:122 like this

#

here is latest log of the proxy

royal hawk
dawn umbra
#

i did

#

both of them

#

first tried the 1024

#

and then the -1

royal hawk
#

And you restarted the proxy between doing that?

dawn umbra
#

yes

royal hawk
# dawn umbra yes

God damnit I was half asleep, I tried to account for using -1 but I was probably too quick. Try 2048 in both

dawn umbra
#

oof

royal hawk
#

No; on second thought I apparently didn’t account for -1

#

I just wanted it to be readable in this case

#

I’ll fix it with the next upstream update

dawn umbra
#

if you can change that during now and tomorrow same time or till friday when i do have time

#

i will change this shit

#

cause atm i do not have enough time to test

royal hawk
#

I mean it still works

#

You just have to use a valid number not -1

dawn umbra
#

ok

royal hawk
#

like ten thousand or even more if you want

dawn umbra
#

when will upstream update be released tho

#

if there is an eta

royal hawk
#

Why would you want to disable this in the first place entirely

dawn umbra
#

so that account for each player or for 10-20 players at once

royal hawk
#

this limit is per connection
Not per proxy

dawn umbra
#

cause that only happens when i join with modded client

royal hawk
#

Yes and the modded clients registers X number of mods

dawn umbra
#

ya

royal hawk
#

It’s a finite number of mods not infinite

#

And I’ve never seen a modpack with over 400 channels

dawn umbra
#

like the ftb modpack ftb revelation has about 200 mods

#

but some mods register 5-6 channels

#

probs more

royal hawk
#

Then set it to 4096 max channels per player

dawn umbra
#

cool

#

will try tomorrow then

royal hawk
#

The limit you can set it to is the positive integer limit in Java

#

But you won’t ever reach that

#

I assume you’re out of memory at around 300k channels

dawn umbra
#

i will try it tomorrow

#

if you can help me with plugins similar to what i use in this waterfall to be able to use it on velocity then i will move over to velocity

#

rather than using waterfall again

royal hawk
#

Velocity has a hard limit of 1024 channels and I’ve never seen a modpack break that record

dawn umbra
#

well it just broke it when i tried it

#

this is all the plugins i use atm

#

IK half of them does work with velocity

royal hawk
#

I should add that you don’t need to change the name limit

#

Legacy fabric is the only offender I’ve ever seen to exceed the name limit

dawn umbra
#

oof

#

welp i will test tomorrow if i have the time

#

but if i can't do it tomorrow afternoon i will do it friday when i am off cause IK i am gonna be tired when i get home afterwork again

#

i will ping you with an update if you would accept me to ping you.

royal hawk
#

I’ll read this channel anyway but idc

dawn umbra
#

👍

tidal musk
#

am trying to setup waterfall with two servers but i cant even see the motd and stuff and cant connect

#

i have two backends running and a waterfall server backend1 is on port 25565 backend2 on 25566 and the waterfall on 25567 all the ports are forwarded

odd hatch
#

Whats the error

royal hawk
royal hawk
royal hawk
#

Are you using docker?

tidal musk
tidal musk
royal hawk
#

any hosting panel type of software?

tidal musk
#

no

odd hatch
#

oh ok

#

I've got a pretty good guess

royal hawk
#

Or just bare using systemd or another demonizing service

odd hatch
#

Is your waterfall server listening on the right interface

royal hawk
#

0.0.0.0 ideally

tidal musk
odd hatch
#

Can you "connect" to your network from the server hosting it?

tidal musk
#

ye

royal hawk
#

Your iptables are in order?

tidal musk
#

nothing like ufw or iptables on the server

royal hawk
#

port/tcp established, new needs to be allowed

#

The waterfall log please then

#

.paste

marble kelpBOT
royal hawk
#

Really don’t know in this case

#

In any case what does iptables -S give you

#

even if you say you don’t use it some system services use it for forwarding anyway

tidal musk
#

oh

#

i mistyped

#

i put 25577

#

instead of 25567

royal hawk
tidal musk
#

im having issues setting up a queue system

#

they tell me the issue isnt with the queue

#

so if the backend reaches the cap then someone gets placed in the queue and someone on the backend leaves the people in the queue cant connect they get "Connection failed. Please try again or contract an administrator."

novel dome
#

should I use waterfall or velocity

surreal stream
eager zephyr
#

I'm getting this error when trying to connect to Lobby. We are using Ptero. (Everything is set correctly in bungeecord and spigot.yml)
Error: -> UpstreamBridge has disconnected

reef fulcrum
#

all that tells us is that the client disconnected

eager zephyr
#

Yeah I know. But it doesn't get thru bungeecord.

reef fulcrum
#

check client logs, etc; If that's all the info you can give us, all I can tell you is what it means

eager zephyr
#

I mean what else can I give lol other than that, that player doesn't get thru bungeecord otherwise it would be connecting to Lobby-1 and therefor it will say in Lobby console anything about this

tidal bolt
#

point me in the right direction if this is not the place... i am trying to set up a forge server in my waterfall. im getting stuck at actually connecting to the server long enough to be sent to the forge server..

wraith blaze
#

sometimes i´m getting this warning what it means?
[23:40:59 WARN]: io.netty.channel.StacklessClosedChannelException [23:40:59 WARN]: at io.netty.channel.AbstractChannel.close(ChannelPromise)(Unknown Source)

reef fulcrum
#

The connection was closed unexpectedly

fallow gate
#

Hey everyone, I've noticed lately that Waterfall is using insane amounts of RAM as does Paper. Now, this all started since we wiped our node (we co-locate our hardware at a hetzner datacenter) and reinstalled pterodactyl and all our servers. We did this to change the way Wings was ran (wings was ran in a screen before, which limits performance in our experience) and now it doesn't. However, since this, our RAM usage has increased by 60-70% across our entire network. We have 128GB of memory so it's not a huge deal but when the network is under load I'm worried Waterfall/Paper (but mostly worried about waterfall as if it goes down the whole network does too) will eat up all it's RAM and crash. Any ideas as to why it would be doing this?

twin solstice
#

(wings was ran in a screen before, which limits performance in our experience)

#

lol

#

what are your startup flags

young rampart
#

If you are using the recommended flags that is completely normal behaviour, the flags make sure that Java reserves the RAM you give it completely when starting the server, that is why you might see it has increased RAM usage, while it's really just reserved RAM

fallow gate
# twin solstice what are your startup flags

For waterfall: java -Xms128M -Xmx2080M -jar waterfall-1.17-454.jar
For Paper: java -Xms128M -Xmx9G -XX:+UseG1GC -XX:+ParallelRefProcEnabled -XX:MaxGCPauseMillis=200 -XX:+UnlockExperimentalVMOptions -XX:+DisableExplicitGC -XX:+AlwaysPreTouch -XX:G1HeapWastePercent=5 -XX:G1MixedGCCountTarget=4 -XX:G1MixedGCLiveThresholdPercent=90 -XX:G1RSetUpdatingPauseTimePercent=5 -XX:SurvivorRatio=32 -XX:+PerfDisableSharedMem -XX:MaxTenuringThreshold=1 -XX:G1NewSizePercent=30 -XX:G1MaxNewSizePercent=40 -XX:G1HeapRegionSize=8M -XX:G1ReservePercent=20 -XX:InitiatingHeapOccupancyPercent=15 -Dusing.aikars.flags=https://mcflags.emc.gs -Daikars.new.flags=true -jar server.jar nogui

twin solstice
#

both of those are wrong

#

xms & xmx should always be equal

#

you should also run wings in a systemd service that pterodactyl provides in the documentation

fallow gate
# twin solstice both of those are wrong

The ones for Waterfall were generated by pterodactyl and haven't been changed. The ones for paper were generated by https://blog.airplane.gg/aikar-flags/ then my systems admin changes the xms flag and I do not know why. I assume to make it match the default ptero flags? Also, that is how wings runs now (to my knowledge). In regards to Pterodactyl's documentation, I've never read it as I didn't install Pterodactyl. That's what I have a systems administrator for. Is there any reason to change the xms flag back? If Ptero by default doesn't have them symmetrical, it hardly seems necessary.

reef fulcrum
#

.flags

marble kelpBOT
reef fulcrum
#

I'd suggest reading that, do note it takes a moment to load due to aikar being dum and using wordpress...

rugged basin
#

your sysadmin is dum

reef fulcrum
#

Memory usage is a complex topic and it's pretty hard to say why memory usage changed between X and Y setups, all it takes is for a speed difference in the CPU and the garbage collector can decide to act entirely differently

fallow gate
reef fulcrum
#

Not saying it isn't...

fallow gate
twin solstice
#

correct

young rampart
#

the wait is worth it

fallow gate
#

So this documentation makes no sense to me. It says don't use all your memory but what does that mean? All of the servers technically have access to 128GB of RAM. Does it mean don't allocate all 128 gigs or don't allocate all of the 16 I have dedicated to that specific server within the machine? Also, it doesn't inspire much faith in things that the last time this post was updated was nearly 3 and a half years ago... and it says if I use more than 12 gigs of memory I need to change some flags? But in the past when I have allocated more than 10 gigs of ram to a Paper/Waterfall server using Aikar flags (default Ptero ones don't) it just causes the server to crash after a few hours of people playing. And on top of that, this post doesn't touch on Waterfall or give any guidance for it. It doesn't even mention bungeecord. Does waterfall need to have it's flags changed? This article leaves me with more questions than answers.

young rampart
#

The recommended flags for waterfall are pinned in this channel

#

Don't use all your memory means don't start Java with all of the memory the system it runs on has available, because it always needs a certain overhead

rugged basin
#

it doesn't matter that it hasn't been updated in 3 years, nothing has changed

reef fulcrum
#

I mean, the reason why I linked the flags was to point out why to do Xmx == Xms

#

What's the point in saying that java can use X amount of ram if you don't want it using that much

#

the recommended flag tweaks for 12G+, I don't recommend using, tbqh

#

Also, ptero is misconfigured by default, basically, advise setting the limits on docker containers to 0 (or just not setting them) and set Xmx/s manually

fallow gate
#

So, if I'm understanding correctly, these flags mostly just enable parallel GC correct?

reef fulcrum
#

G1GC, which is the default now, but, those flags specifically tune the collector more towards the behavior of minecraft

fallow gate
#

Is it not possible to use Shenandoah or ZGC?

reef fulcrum
#

We don't recommend them for MC

fallow gate
#

Is there a reason for that?

reef fulcrum
#

they're more designed for large heaps with reasonable throughput, not small heaps with relatively high throughput

fallow gate
#

Ah I see. I was just curious as ZGC reports to be much faster.

#

Do you recommend a particular JDK distribution or a particular OpenJDK developer?

young rampart
#

Microsoft, Corretto, Temurin, doesn't really matter that much

fallow gate
#

Really? I had always heard that JDKs can vary wildly from developer to developer.

#

Perhaps I was misinformed.

young rampart
#

I've mainly used AdoptOpenJDK in the past, which is Temurin now, but because they still have no apt repo I recently switched to Corretto without any noticeable difference

reef fulcrum
#

It's all mostly just repackages of openjdk

#

some of them have some tweaks for them, e.g. some of them ship builds for m1, etc; but, for the most part they're all the same

fallow gate
#

Interesting. I actually noticed a performance dip when we used OpenJDK. We recently switched to Oracles JDK and the performance increase was very noticable.

reef fulcrum
#

back in like the 1.8 era it did matter where oracles JRE offered better performance but these days they're all practically the same

fallow gate
#

I was wondering why that was the case.

#

Why some servers seemed to hate OpenJDK and others didn't seem to care.

#

At any rate though, looking at our skyblock server, it's using 8 of it's 9 gigs (I recently lowered it from 16 due to the before mentioned crashing issue with ram numbers above 10 gigs) but it isn't lagging. I'm assuming it's not because this ram is occupied but because it's been reserved?

reef fulcrum
#

You'd need to attach a profiler to the server to see what's going on

#

the JVM has never been great at freeing ram, if your servers performing better now it might even be that it's just generating more junk in the heap faster cos it's performing better, etc; not to piss in the wind too much there

fallow gate
reef fulcrum
#

Generally you don't care about heap dumps until you're actually looking at what is in the heap, generally you just wanna see if it's behaving "healthily"

fallow gate
#

I might have forgot to turn off the ping. If I did I'm sorry.

reef fulcrum
#

I mean, am a yourkit person, but, for what you're doing, visualvm would work (might need a plugin if you wanna see what old/new gen is doing instead of just used/free), or jprofiler would work too

fallow gate
#

Actually, because I can't find any other information on it, what does the whole old gen and new gen thing mean? ELI5

reef fulcrum
#

I mean, in this case you just wanna see what the memory is doing, not really look at the "profling" aspects, just, that's what the tools are often referred to as; big thing is seeing how the memory is "spiking", with g1 it should generally look a bit like a saw blade, ideally you want the pits to remain roughly around the same place

#

G1 is generational, new is "newly allocated objects", stuff as it sits around progresses into "old gen" which is collected less often

fallow gate
reef fulcrum
#

No

fallow gate
#

How about Grafana?

reef fulcrum
#

unless it has a java agent thing, the memory of the java process itself doesn't represent too much given the whole memory regions of java

fallow gate
#

So monitoring the whole machines ram usage won't help me... welp... looks like I'm learning how "yourkit" works lol

fallow gate
reef fulcrum
#

leave some for the OS, make sure to leave some for java itself too

#

.heap

marble kelpBOT
#

-Xmx controls the amount of heap memory assigned to the JVM, this does not include memory used for java itself and the native memory used by java itself, as well as other libraries such as netty (for networking) and SQLite.

fallow gate
#

Ofc leave enough for Ubuntu to run but does that mean then just don't use all the total system memory?

reef fulcrum
#

Xmx doesn't cover the full memory usage of the JVM

#

especially in ptero it's important that you leave enough room for overheads

fallow gate
#

How much overhead though? 8 Gigs? 20? I can spare quite a bit lol

reef fulcrum
#

that's more pointed out for ptero users given that ptero sets the memory limit for the container to whatever the Xmx thing is for, it does give a bit of a buffer but not always enough

#

Easier to just set the container limit to 0 and set Xmx/s manually

fallow gate
#

If I set the container limit to 0 it refuses to launch the server.jar for some reason.

#

It just gives an out of memory error then crashes

reef fulcrum
#

there should be a way to tell ptero to not limit the containers memory

#

I don't use it so not super familiar with it

fallow gate
#

It says to set the container memory limit to 0 for unlimited but it just doesn't work for some reason

mellow rock
#

try -1

pale valve
#

I am running Hexxit II with waterfall, and I was wondering how do I go about fixing the issue where it shows an X on the server list? I.e. before Waterfall it showed mod count.

dawn umbra
#

we still getting this message

marble kelpBOT
#

enabling ping-passthrough might fix it but it might break

dawn umbra
#

and i have changed the value

marble kelpBOT
#

*break with modded handshakes I guess

dawn umbra
#

in the waterfall.yml

marble kelpBOT
#

well increase it

#

or remove whatever is spamming channels

dawn umbra
#

i allready have it on 10240

#

can't

#

cause it is a modpack and i can't remove mods from server

marble kelpBOT
#

of course you can? you just don't want to lmao

dawn umbra
#

dude if i remove it fucks the pack up

#

sorry for the language

marble kelpBOT
#

so?

pale valve
#

Yeah you cant really remove mods from modpacks, as both sides need it.

dawn umbra
#

and i do not know what mod it is tho

pale valve
#

What modpack are you running, @dawn umbra ?

dawn umbra
#

so i am not gonna do a binarry scan

marble kelpBOT
#

either get teh mod author to fix their shit or use a higher count or don't use a (broken) proxy

dawn umbra
#

FTB revelation

pale valve
#

Glad I haven't tried revelations on a bungee setup yet xD

marble kelpBOT
#

@pale valve yes, you would have to remove it from the modpack and provide a modified version of it

dawn umbra
#

it worked and it just didn't work after a proxy update

pale valve
#

Ah, try downgrading your proxy?

#

If it worked before

reef fulcrum
#

er, what option do you have set that high?

marble kelpBOT
#

inb4 they set the name limit

dawn umbra
#

plugin_channel_name_limit: 10240

reef fulcrum
#

yes

marble kelpBOT
#

🤦‍

pale valve
#

Raise it then, I assume

marble kelpBOT
#

no

#

...

#

use the correct config option

reef fulcrum
#

That's the maximum length of the name of the channel

#

Not the maximum number of channels

pale valve
#

Ah

pale valve
marble kelpBOT
#

yes

pale valve
#

Doesn't ping-passthrough mess up player counts, etc?

#

If the player count is spread cross multiple hexxit ii servers

marble kelpBOT
#

yes, that's the goal of the option

pale valve
#

Hmm, thats unfortunate.

marble kelpBOT
#

use a server-side plugin/mod to display the global count in the ping response if it works I guess

pale valve
#

Hmm, okay. Ill look into that.

dawn umbra
#

here is the channel error i get

reef fulcrum
#

and your config?

dawn umbra
#

here is proxy config

reef fulcrum
#

the waterfall config given that that is the one we've been talking about

dawn umbra
#

or do you want the waterfall.yml

#

cause i can't seem to find what you been kinda talking about

reef fulcrum
#

yes

dawn umbra
#

wait are you saying i need to recreate my config file

reef fulcrum
#

no?

dawn umbra
#

here is the yml

reef fulcrum
#

registered_plugin_channels_limit

dawn umbra
#

yes

#

and

#

that is set to 10240

reef fulcrum
#

No

#

you've been tweaking the wrong one

dawn umbra
#

oh fuck i seem to have hard time reading today

#

ok it seemed to work

errant pagoda
#

is it possible for waterfall or bungee to "block" datapacks? a scripter is complaining about datapacks not working but i doubt its from it

reef fulcrum
#

no

#

the proxy has nothing to do with datapacks itself, that's all on the server

errant pagoda
#

alright,cheers

pale valve
#

My players keep getting this error, randomly, not all players. Is this related to waterfall?

quartz mauve
#

Damm that got nukes fast paperOhhh

quartz mauve
pale valve
#

No errors.

tulip surgeBOT
pale valve
#

Jon's Exclusives' remote config's mode is invalid happens for all logins btw

royal hawk
pale valve
#

The issue happens on all logins for that message, I think its irrelevent

#

Even for successful logins.

#

[21:39:50 INFO]: This server is running Mohist version 1.12.2-262 (MC: 1.12.2) (Implementing API version 1.12.2-R0.1-SNAPSHOT, Forge version 14.23.5.2855)

royal hawk
#

I should say that forge-spigot hybrids are very unsupported whereever you go

fallow gate
#

Is anybody around who knows IPTables and can help me out? PLease ping me if you can 😄

fallow gate
# loud tendon What do u want to achieve?

Well, I'm looking to prevent an IP range from connecting to the node. There is this guy who is managing to dodge IP bans (maybe through some vpn that can get passed anti-vpn systems idk) and he is just kinda changing the last number of his IP. So I just want to ban the entire IP range. afaik this can only be done by denying the connection from the ip range with iptables. I just don't know how.

loud tendon
fallow gate
#

Ok so I would want to do the first one then. Thanks so much!

#

You just saved me like hours of headache trying to find it in iptables documentation.

vast solstice
#
[16:14:02] [Netty Boss IO Thread #1/INFO]: Listening on /0.0.0.0:25565
[16:19:53] [main/ERROR]: Exception in thread "main" java.io.IOError: java.io.IOException: Erreur d'entrée/sortie
[16:19:53] [main/ERROR]:     at org.jline.keymap.BindingReader.readCharacter(BindingReader.java:143)
[16:19:53] [main/ERROR]:     at org.jline.keymap.BindingReader.readBinding(BindingReader.java:110)
[16:19:53] [main/ERROR]:     at org.jline.keymap.BindingReader.readBinding(BindingReader.java:61)
[16:19:53] [main/ERROR]:     at org.jline.reader.impl.LineReaderImpl.doReadBinding(LineReaderImpl.java:848)
[16:19:53] [main/ERROR]:     at org.jline.reader.impl.LineReaderImpl.readBinding(LineReaderImpl.java:868)
[16:19:53] [main/ERROR]:     at org.jline.reader.impl.LineReaderImpl.readLine(LineReaderImpl.java:575)
[16:19:53] [main/ERROR]:     at org.jline.reader.impl.LineReaderImpl.readLine(LineReaderImpl.java:418)
[16:19:53] [main/ERROR]:     at net.minecrell.terminalconsole.SimpleTerminalConsole.readCommands(SimpleTerminalConsole.java:158)
[16:19:53] [main/ERROR]:     at net.minecrell.terminalconsole.SimpleTerminalConsole.start(SimpleTerminalConsole.java:141)
[16:19:53] [main/ERROR]:     at net.md_5.bungee.BungeeCordLauncher.main(BungeeCordLauncher.java:72)
[16:19:53] [main/ERROR]:     at net.md_5.bungee.Bootstrap.main(Bootstrap.java:15)
#

wath is me probleme ?

reef fulcrum
#

generally means that you didn't run the proxy in a proper console or you have a janky console

reef fulcrum
#

don't mention

#

how are you running the jar?

vast solstice
#

it's the last waterfall the script to turn on the server is still the same i don't understand

reef fulcrum
#

okay

#

and how are you running that script?

reef fulcrum
#

not too sure in that case, screen has been kinda janky over the years, but, otherwise 🤷‍♂️

vast solstice
#

yesterday it worked

oak crane
#

did you change anything to the script ?

royal hawk
vast solstice
#

server update java 16 to 17

#

the problem is path java xD

oak crane
#

btw in your 8th line i'm pretty sure adding nogui as an argument is usless since waterfall doesnt have a gui

#

correct me if im wrong tho

royal hawk
#

That doesn’t fix the underlying issue. You’re not specifying an env to run it in. That it works on an older java version is coincidence

vast solstice
#

most of my 14 servers use java 8 and waterfall needs java 16 for geyserv

#

If not, thank you, I have no more problems.

oak crane
#

how can i do it so when my players reconnect they go in the hub server instead of the server they were in?

oak crane
royal hawk
oak crane
#

it works now, had to change force_default_server: false to true in my config.yml

#

yeah didnt notice it was on false by default

steel surge
#

Was about to say that, also hi minecon

oak crane
#

lmfao what

#

since when are you here

oak crane
vast solstice
#

i go test XD

#

so normally kicks and bans or the server stops or restarts come back in the server default

tidal musk
#

Hi is it possible to connect waterfall to another waterfall server?

reef fulcrum
#

no

royal hawk
#

In theory; yes; but not without extensive modifications to waterfall

#

And it goes without saying that we won’t help with that

prisma hedge
#

what java version i should use for waterfall?

severe badger
#

Java 17 will work fine

narrow flower
#

it keeps thinking that everyone's ip is 127.0.0.1, causing this: anyone know a fix?

lost river
narrow flower
#

enable ip forwarding on your proxy
how?
and bungeecord in the spigot.yml of the other servers.
done

tidal musk
#

ip_forward: true in bungee.yml

narrow flower
#

ig it's waterfall.yml for waterfall

#

nope ok

tidal musk
#

config.yml* my bad

narrow flower
#

k

#

same error

lost river
#

did you restart both the proxy and the backend server?

narrow flower
#

i rebooted the proxy

#

havent done the 2 servers yet

lost river
#

you also need to restart them so the bungeecord setting takes effect

narrow flower
#

both servers have now done

#

still same error

#

strange

lost river
#

what plugin do you use for that message?

narrow flower
#

wdym

#

the proxy has no plugins

lost river
#

oh, sorry, ip forwarding and ips are not related to that error message

narrow flower
#

ok

lost river
#

you are trying to connect to the proxy twice

narrow flower
#

no

lost river
#

with the same account

narrow flower
#

no im not

#

the server sees everyone as the same ip:
[15:58:07 INFO]: [/127.0.0.1:46135] <-> InitialHandler has connected
[15:58:07 INFO]: [Deveroonie] disconnected with: Kicked whilst connecting to lobby: You are already connected to this proxy!

lost river
#

if you do glist in the proxy console, is there anyone online?

narrow flower
#

glist
[16:00:56 INFO]: [vanilla] (0):
[16:00:56 INFO]: [smp] (0):
[16:00:56 INFO]: [lobby] (0):
[16:00:56 INFO]: Total players online: 0

lost river
#

ok, strange

tidal musk
#

Are you sure lobby doesn't point to the same ip as your bungee?

reef fulcrum
#

inb4 you have waterfall tryna connect to itself

#

shut up noah

#

just cos you can tappey key faster, you know

narrow flower
#

when i join the console goes like this

[16:01:34 INFO]: [Deveroonie|/ee.ee.ee.ee:54764] <-> ServerConnector [lobby] has connected
[16:01:34 INFO]: [/127.0.0.1:46339] <-> InitialHandler has connected
[16:01:35 INFO]: [Deveroonie] disconnected with: Kicked whilst connecting to lobby: You are already connected to this proxy!
[16:01:35 INFO]: [/81.109.31.135:54764|Deveroonie] -> UpstreamBridge has disconnected
[16:01:35 INFO]: [Deveroonie|/81.109.31.135:54764] <-> ServerConnector [lobby] has disconnected
narrow flower
reef fulcrum
#

No, bungee is a proxy

narrow flower
#

ah shit

#

on the hub how do i send my players to another server then

lost river
#

there is an entry in the config.yml for that

narrow flower
#

ok

lost river
#

there is a list of server names that the proxy tries to connect you to on login

narrow flower
#

should i give the proxy : 25565

tidal musk
#

If you want players to be able to connect without specifying a port, yes

reef fulcrum
#

there are plugins which you can install which will offer "pretty" ways to teleport across servers, as well as the built in /server command for players

narrow flower
#

ok

#

"pretty" ways
shit like NPC's?

lost river
#

or portals

narrow flower
#

ok

#

all the servers are up, now i get this:

#

the ports are open and the servers are up

#

i direct connected to the hub server and got this:

#

can anyone help me fixit#

reef fulcrum
#

the proxy failed to connect to the server, check the logs, etc

#

boils down to your setup and how you're running the servers, etc

narrow flower
#

logs dont suggest any issues

narrow flower
reef fulcrum
#

see ptero's documentation on setting up a network

narrow flower
#

should be the same as running it all on localhost

reef fulcrum
#

No, because the containers localhost is on its own network

narrow flower
#

wdym

reef fulcrum
#

it's a container