How do bots effect a server?

[dexter610]

Im cuorious to know if bots take up the same ammout of resources as a real player? (aside from upstream)

In other words, is adding a bunch of bots a good way of test what your server can handle?

[BehaartesEtwas]

not at all. bots use the cpu completly different. i have seen servers behaving better and others behaving worse with bots compared to human players. don't use them for testing!

[Zykomaniac]

From what I've seen on bots cpu usage it sometimes uses 4 times more cpu per player on Q6600 and if heavy firefight 10-20Times more cpu per player
and when spawning when the bots are given their waypoints/navigation points 8-12 Times more varying on map. they also consume more memory although I cant give an answer on how much more memory it consumes. However they use less bandwidth download, but the players must upload the info they get from the server regarding the bots so the upload bandwidth stays the same or higher.
Also results may vary.

For me, different difficulty levels also got a large impact on how much cpu usage they suck from the server, I tested on difficulty level medium(3 I think?)

This was testet at
server at 1000fps
tickrate 100
rate 25000
cl_updaterate 100
cl_cmdrate 100
comparison was
6 players
and one test with 6 bots

also bots at 100tick and fps at 1000fps can easily make the server being noticed as laggy/jumpy

[BehaartesEtwas]

maybe true, but you cannot conclude from the cpu load to e.g. stability of the fps. I've seen servers running very stable (w.r. to the fps) with bots on them, but as soon as human players test it, it completely collapses. of course if you but so many bots on the server that it has 100% cpu load, you will have lags (which might vanish for human players).

just don't do any tests with bots unless you intend to run bots on the server.