[cvsnt] Re: Stress Tests results for CVSNT/CVS/Subversion

[Tony Hoyle]

Rahul Bhargava wrote:

> With 50 clients pounding on a CVSNT server (2.5.03) running on a Windows 
> 2003 Server machine
> with 1GB RAM, 2GB SWAP, 2xPentium 4 CPUs (2.8GHz Dell server class 
> machine), we saw that after
> about 15 minutes of stress the CVSNT Lock Daemon service would freeze, 
> the CPUs would be maxed out at
> 100%. When the freeze happened, almost always the command `rtag' would 

I've run successful tests with over 1000 clients (the OS gets real slow 
under that kind of load so the cvs processes end up not being the 
bottleneck)...  50 is not a stress test.  50 what happens when an office 
starts up in the morning and everyone runs cvs update... it's routine 
behaviour.

Without seeing your scripts and logs to see what you are actually doing 
I can't even begin to suggest what might be the issue there.

Failure over time doesn't make any sense - each cvs invocation is 
unique.  Are your scripts not waiting until the previous command has 
finished?  If you're not waiting you could well be testing 50,000 
clients not 50.

Tony