UnrealIRCd Forums

Hi everybody,

At the moment I'm busy doing some tests to see how easy it'd be to replace Hybrid7 (my current ircd) with unreal 3.2. So far, so good: things work very well and it turns out that my irc services (Anope) really agree with Unreal (in comparison with Hybrid7).

One thing puzzles me: The amount of users. In a thread on the Anope forum I read a few comments which basicly boiled down to "If you have a massive user amount then Hybrid7 may be better capable of handleing the load, but for regular networks Unreal should do just fine.".

Now, I don't think my network is big enough to even worry about this issue but I'd like to be prepared: At which user amounts should I start worrying about the load and what other signals would point me to possible problems ?

My first guess would be load average, memory usage and basic load of the ircd, but are there any more hints ?

Thanks in advance for all the input!

As has been posted many times, it's quite subjective; it depends not only on how many users you have, but what those users are doing.

If you're interested in how many users can be connected to an UnrealIRCd server simultaneously: on *NIX it defaults to 1024 (but can be tweaked, see the FAQ for more on this), while the Win32 build can (supposedly) handle up to 4096 (dependent on OS version).

Ron2K wrote:If you're interested in how many users can be connected to an UnrealIRCd server simultaneously: on *NIX it defaults to 1024 (but can be tweaked, see the FAQ for more on this), while the Win32 build can (supposedly) handle up to 4096 (dependent on OS version).

I was aware of the limitation per server but in this aspect they were doing a very broad comparison between the way the several ircd's handled the load/stress of the whole network.

It basicly boiled down to one ircd doing more internal checks than the other. For example; in Unreal its possible to scan the contents of whats being said, and as an automatic result this could increase the load on the server.

But I do see your point: it seems I allowed myself to start comparing apples and oranges without even realising it at first (since, when looking at my example, you can also turn that stuff off thus removing this difference).

Thanks for your response.