On Mon, 2003-11-24 at 14:36, Geoffrey Young wrote:
> in the spirit of only doing what you have to, we were 
> kicking around the idea of killing the largest interpreter in the pool 
> rather than bringing down the entire process (which replacing in an mpm such 
> as worker might add significant overhead).

That makes sense, but also sounds like it could get rather complicated.
I'd like to get the prefork working before we get mired in thread stuff.

> I was going to 
> look into doing something similar for 2.0 by twiddlins some of apache's 
> private bits to get the same effect.  we'll see how it goes :)

That would certainly be very useful, and not just in SizeLimit.  I've
used child_terminate() when I have to load some large and rarely-used
module into memory.

> IIRC, it was something about RSS not being accurate

That sounds likely.  The size calculations for each platform are
contributed by different people, and I only have Linux available for
testing.  The good thing is, the size is only a few lines of code, and
easy to replace.

- Perrin


---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]

Reply via email to