Re^3: Does IO::Select work? Solution

[zentara]

in reply to Re^2: Does IO::Select work? Solution
in thread Does IO::Select work? Anywhere?

So that 100% cpu usage, is where you think the kernel needs tweaking? It's an odd 100%, because it dosn't seem to slow down my other processes. Maybe the kernel processing reporting is deceiving, just like the problem with memory gains, where even though ps says memory is being held by a process, the memory gain is just in the swap area, or some technical detail, which isn't real memory.

Is this 100% cpu usage really 100%? If so, how come all my other apps are still as responsive as ever?

[tye]

Just change the time-out, probably from 0 to "infinity" (which is likely represented by -1).

[zentara]

Tye, I didn't think you cared about anything I wrote!. :-) I tried -1 for infinity, and still 100% cpu. But I had the audacity to undefine infinity and it does the right thing. CPU usage is minimal.

#Change:
#my ($new_readable) = IO::Select->select($readable_handles, undef, und
+ef, 0 );
# to
my ($new_readable) = IO::Select->select($readable_handles, undef, unde
+f, undef );

and the server in Re: Does IO::Select work? Solution is the Holy Grail after all. No sysctl.ph needed. :-) Minimum cpu usage.