Messages in this thread | | | From | Peter Steiner <> | Subject | Re: [patch] 2.2.9_andrea-VM1.gz | Date | Tue, 8 Jun 1999 22:45:39 +0200 (CEST) |
| |
> >> I forgot that ext2/truncate is still buggy in 2.2.9. > > > >It didn't happen with plain 2.2.9... At least I didn't notice it. > > Yes of course, my patch triggers the bug but it's a bug in ext2fs (it's > present also in ufs and minix btw) and it's just been fixed in 2.3.x > infact.
It works now, thanks.
> >>--8<-- > > > >Maybe an average bucketsize of 8 is too high? This is sufficient with 64MB > >since the buffer cache doesn't grow so much anyway. Try reducing it to 4 > >(The original hashfn has an average bucketsize of 1). > > Good idea, I'll set the distribution to 1, I don't want to lose because I > am more conservative in memory.
Could someone please tell me the typical size of the buffer cache in big real world servers?
An average bucketsize of 1 is too much for the average desktop but it may indeed help for servers. It seems there should be a config option:
"[*] Optimize for big buffer caches" --> avg bucket zize = 1 else avg bucket size = 4
And for 386/486 there should be a shift/add hashfn. On a 486 it takes 40 cycles for one multiplication and it would be nice to still support these old PCs.
Peter -- _ x ___ / \_/_\_ /,--' p.steiner@t-online.de (Peter Steiner) \/>'~~~~// \_____/ signature V0.2 alpha
- 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/
| |