XRV-Webix
Member
- May 2, 2016
- 111
- 55
- 64
Hello @maxst,
I had more than 3000 connected at the same time on the "backend" server.
Hard to tell how many were passing thru each of the "frontend" because i wasn't monitoring the connections.
About informing the teamspeak on the weblist, is not a such good idea... You'll have duplicates on the weblist, or, you will force all clients connecting from the weblist to use the same "frontend" server. In my case, i didn't advertised the server at all.
But, the good thing of using the iptables to redirect the traffic, is that you can analyse the traffic on each "frontend" server and "re-analyse" it again on the "backend", like i did.
With the "frontend" servers with no anti-ddos protection from the provider, the attacks targeted to the TS server just maked crash 1 or 2 "frontends" where some TS users are dropped and get back online right away using another "frontend" server.
Note: To do a setup like this, you will always need the "frontends" have 2 network cards
I had more than 3000 connected at the same time on the "backend" server.
Hard to tell how many were passing thru each of the "frontend" because i wasn't monitoring the connections.
About informing the teamspeak on the weblist, is not a such good idea... You'll have duplicates on the weblist, or, you will force all clients connecting from the weblist to use the same "frontend" server. In my case, i didn't advertised the server at all.
But, the good thing of using the iptables to redirect the traffic, is that you can analyse the traffic on each "frontend" server and "re-analyse" it again on the "backend", like i did.
With the "frontend" servers with no anti-ddos protection from the provider, the attacks targeted to the TS server just maked crash 1 or 2 "frontends" where some TS users are dropped and get back online right away using another "frontend" server.
Note: To do a setup like this, you will always need the "frontends" have 2 network cards