Hi,

I need some advice to get myself out of a stupid situation I put myself 
into.

I have a server that I was testing with for resources, etc.

I was testing the effect of sysctl kern.maxproc=xxx to see if I could 
figure out the memory usage of httpd until I get my additional memory in 
the server as it's in order and I don't have it yet.

But now, I have more process running then the kern.maxproc value and I 
can't even to pkill httpd for example as it gets

sh: cannot fork - try again.

or in a different shell:

ksh: cannot fork - try again

So, how can I either reboot the server, or kill that process, etc?

I guess I have to drive to the site to reboot the server manually now? I 
hope not, but that may be the only thing possible now.

Why I was doing this? Well, because the server was crashing with

extend_alloc_supregion: can't allocated region

and

pf_test: pf_get_frag returned NULL

So, I was trying to find a temporary stage where I would be able to 
maximize the outcome, but at the same time avoid the crash.

I still don't know what the above message mean, but anyway, it look like 
memory get low and then swap start to be use and very shortly after 
that, all usage go out to the roof and then crash.

Any possibility to do this in my shell.

I have two opne, one using ksh and the second using sh.

But both don't let me fork anything.

Any advise?

Thanks

Daniel