On 19 Mar 2000 15:41:30 +0100, you wrote:
>Den 18-Mar-00 01:24:49 skrev Linda Walsh følgende om "Re: On the issue of low memory situations":
>
>> I haven't read through this whole thread, so this may have been
>> suggested, but why not have a new signal "SIGNMEM". Can't be caught but
>> can be ignored. Default is to take the signal and terminate the program
>> that faulted. If ignored, put process to sleep until the memory request
>> can be satisfied. Then something like 'X' or apache could ignore, while
>> 'gcc' would just die.
>
> It would be much more useful to make it possible for SIGNMEM to be
>caught. This would be similiar to low memory handlers as seen on AmigaOS.
>Programs could then register a signal handler for SIGNMEM and free up
>memory that isn't needed when the signal handler is caught. Some programs
>cache reconstructable data for speed reasons and some daemons keep idle
>children running to save fork()/exit() overhead.
Yes. Apache, for example, could reduce the number of "spare" children,
or even the number of concurrent requests handled, making quite a big
difference. (Ideal in a load-balanced server farm - this would
automatically transfer some load to less busy systems.)
> Just wondering: Linux isn't smart enough to unload unused modules in low
>memory situations, is it?
Sounds like a good idea....
James
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu
Please read the FAQ at http://www.tux.org/lkml/
This archive was generated by hypermail 2b29 : Thu Mar 23 2000 - 21:00:29 EST