Subject: Re: Bad response...
To: Johnny Billquist <bqt@Update.UU.SE>
From: Frederick Bruckman <fredb@immanent.net>
List: current-users
Date: 08/29/2004 12:42:13
In article <Pine.LNX.4.61.0408272057580.18681@psilocybe.update.uu.se>,
bqt@Update.UU.SE (Johnny Billquist) writes:
> On Fri, 27 Aug 2004, Manuel Bouyer wrote:
>
>> You also need to bump execmax and anonmax, otherwise some memory may still
>> be allocated to the data cache while the system is paging.
>
> Manuel, execmax and anonmax is in that list you quoted. :-)
>
> Anyhow, I've twisted all the vm.*min and vm.*max, and actually did get the
> response back into the system. But which parameter did it, or if it was a
> combination of them, I don't know.
I have noticed that raising *vm.execmin* from "5" to "10" makes for a great
improvement with the issue you described on low memory machines.
--
Frederick