Messages in this thread | | | Date | Fri, 27 Mar 1998 20:44:15 -0500 (EST) | From | Adam Heath <> | Subject | Re: 2.1.91 swap performance: jerky. |
| |
On 28 Mar 1998, Linus Torvalds wrote:
First off, I am running 2.1.91. P166MMX. 32mb ram. 64mb swap. The events in question occur when I am running pine, mirror in the background, tcpdump, 2 pppstats, fetchmail, along with 6 bashes.
> In article <351C2E91.5C23733D@pobox.com>, mlord <mlord@pobox.com> wrote: > > > >Tried it. Less dramatic swings in swapping behaviour > >than before, but still leaves more free memory than I'd > >expect.. 4-5MB "free" at all times, despite continuous swapping. > > This is intentional. The background swapping tries to swap out enough > stuff that we're never really low on memory (and that means that it > tries to keep at least one 128kB _contiguous_ area free, for example - > which in turn tends to indicate a few megs of real free memory).
Why? Why do the swap out ahead of time? Then the kernel needs to swap in the swapped out code. I think the swapout should happen in the memory allocation code. Or, the swap out stragety needs to be redone. Too often, I see currently active programs being swapped out, only to have to be immediately swapped back in.
> >Either with or without, interactive response still "feels" okay, > >considering the system load. > > Good. This is actually the major aim of the 2.1.90 code: to keep the > system interactive "feel" the right way, by having a background swapout > that is fairly aggressive but really stays in the background (and it is > designed to be aggressive enough that it should hopefully never get to > the state where the machine gets so tight on free memory that it starts > to feel unresponsive).
Yes, the system is be more responsive than 2.1.91pre1. But still, it is aggravating having things swapped out while they are being used(the bash shells).
Adam
- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.rutgers.edu
| |