On 1 Mar 2000 nbecker@fred.net wrote:
>Thanks for all the interesting replies to my question.
>
>It appears to me that there is a kernel design problem. If all that
>happens when someone creates very large directories is that their
>application runs slow, we could argue that that was just a user
>programming issue. But it appears to me (based on only nonscientific
>observation) that not only does that app slow down. I suspect that we
>are spending a large amount of time in non-preemptable kernel. I say
>this because when I run on my 400MHz i686 I am usually not bothered by
>a second user running some process. In this case the load is very
>noticable.
>
>If I am correct that a lot of time is being spent in the kernel (I'm
>only guessing at this, so tell me if I'm wrong) then it would appear
>to me that this is not just a user space issue - but really a weakness
>in the kernel design. If so perhaps higher priority should be given
>to correcting it.
I think btree's would speed up dir operations, however it is
still a combo userland/kernel problem. Poorly designed apps
using poorly designed directory structures should be rethought
and reprogrammed to do it in a more efficient manner
period. Then it will run much faster on existing kernels, and
even more so if btree's are used.
A lot of OS's suffer from this problem, so it isn't just a Linux
issue.
-- Mike A. Harris Linux advocate Computer Consultant GNU advocate Capslock Consulting Open Source advocateSuspicious Anagram #4: Word: PRESIDENT CLINTON OF THE USA Anagram: TO COPULATE HE FINDS INTERNS
- 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 : Tue Mar 07 2000 - 21:00:10 EST