Server Query Flood Protection too harsh

[Multivit4min]

Hello,

I was messing around with the new implemented flooding protection for my Server Query Library,
at first it seemed to work as intendet, got all 2 to 3 commands the error id 524 which told me to wait 1 Second.

I parsed the message to get the exact time i need to wait and added 50ms additionally just in case

After a few hundred successful whoami commands it suddenly told me to wait for 600 seconds (thats a damn long waiting time)
After those 600 seconds +50 milliseconds ran out I resent the command and got a new Timeout for 301 more seconds
and after those 301seconds +50 milliseconds it just disconnected me from the query without a real message

so first of all one big problem I spotted is the new query timeout of 300 seonds, I automatically send all 200 seconds
a whitespace + newline to the query to not timeout.
So after my understanding a Server Admin is FORCED to set the timeout higher
in order to no run into those problems with Query Timeouts and flooding protection.

Note: I used to test this on my local machine and emptied the query_whitelist.txt
Server Version used: 3.3.1

Wouldnt it be easier for us to handle it by just defering the return of a command?
Hope you consider to tweak that a bit more

[samurai-virgin]

Hello,

I was messing around with the new implemented flooding protection for my Server Query Library,
at first it seemed to work as intendet, got all 2 to 3 commands the error id 524 which told me to wait 1 Second.

I parsed the message to get the exact time i need to wait and added 50ms additionally just in case

After a few hundred successful whoami commands it suddenly told me to wait for 600 seconds (thats a damn long waiting time)
After those 600 seconds +50 milliseconds ran out I resent the command and got a new Timeout for 301 more seconds
and after those 301seconds +50 milliseconds it just disconnected me from the query without a real message

so first of all one big problem I spotted is the new query timeout of 300 seonds, I automatically send all 200 seconds
a whitespace + newline to the query to not timeout.
So after my understanding a Server Admin is FORCED to set the timeout higher
in order to no run into those problems with Query Timeouts and flooding protection.

Note: I used to test this on my local machine and emptied the query_whitelist.txt
Server Version used: 3.3.1

Wouldnt it be easier for us to handle it by just defering the return of a command?
Hope you consider to tweak that a bit more

Uhm, agree that. My bots (ip is whitelisted disconnected every 5 minutes). Need to fix that in 3.4.0

[Multivit4min]

alright the keepalive seem to not cause any problems

1535754481001 'SEND' 'whoami'
1535754481002 'error id=524 msg=client\\sis\\sflooding extra_msg=please\\swait\\s1\\sseconds'
1535754481002 'DEFERING' 1 's >>> total wait' 1050 'ms'
1535754482064 'SEND' 'whoami'
1535754482065 'error id=0 msg=ok'
1535754482066 94
1535754482066 'SEND' 'whoami'
1535754482066 'error id=0 msg=ok'
1535754482067 95
1535754482067 'SEND' 'whoami'
1535754482067 'error id=0 msg=ok'
1535754482067 96
1535754482068 'SEND' 'whoami'
1535754482068 'error id=524 msg=client\\sis\\sflooding extra_msg=please\\swait\\s1\\sseconds'
1535754482068 'DEFERING' 1 's >>> total wait' 1050 'ms'
1535754483122 'SEND' 'whoami'
1535754483123 'error id=0 msg=ok'
1535754483123 97
1535754483124 'SEND' 'whoami'
1535754483124 'error id=0 msg=ok'
1535754483125 98
1535754483125 'SEND' 'whoami'
1535754483125 'error id=524 msg=client\\sis\\sflooding extra_msg=please\\swait\\s600\\sseconds'
1535754483126 'DEFERING' 600 's >>> total wait' 600050 'ms'
1535754733143 'KEEPALIVE \' \\n\''
1535754733146 'error id=524 msg=client\\sis\\sflooding extra_msg=please\\swait\\s351\\sseconds'
1535754733146 'DEFERING' 351 's >>> total wait' 351050 'ms'
1535754983153 'KEEPALIVE \' \\n\''
1535754983157 'error id=524 msg=client\\sis\\sflooding extra_msg=please\\swait\\s101\\sseconds'
1535754983157 'DEFERING' 101 's >>> total wait' 101050 'ms'
1535755084212 'SEND' 'whoami'
1535755084215 'error id=0 msg=ok'
1535755084216 99
1535755084216 'SEND' 'whoami'
1535755084216 'error id=524 msg=client\\sis\\sflooding extra_msg=please\\swait\\s1\\sseconds'
1535755084217 'DEFERING' 1 's >>> total wait' 1050 'ms'

those are the lasst few lines i logged, i am currently using 250s for the keepalive interval
but 600seconds are just too insane
(the numbers just count how much whoami commands have been successfully retrieved)

[numma_cway]

I believe OP did not understand flood protection. Flood protection comes from serverinstance_serverquery_flood_*. TeamSpeak's point is that anonymous users and serveradmin should have the same flood limitations. Whitelist should be used, which is pointless for dynamic IPs.

Query timeout is not related to flooding.

Uhm, agree that. My bots (ip is whitelisted disconnected every 5 minutes). Need to fix that in 3.4.0

Whitelisted bots do not disconnect.

[Multivit4min]

but why does it still ban me for 600 seconds?
The Query tells me to wait 1 second, this is exactly what I do and still get banned by flooding

[mmuenchow]

I made a post about how the new query system and the changes to the flooding. You can find it here: https://forum.teamspeak.com/threads/...w-Query-System
If there is anything unclear please post your question in either thread, and i will try to answer them.

On how to solve you problem: respect the flood error, and reduce the rate of requests send. In order to prevent the flood ban(the 600s increase) to hit you, never cause more then one flood warning per second.