Messages in this thread | | | Date | Thu, 29 Jul 2004 13:31:41 +1000 | From | Nick Piggin <> | Subject | Re: dentry cache leak? Re: rsync out of memory 2.6.8-rc2 |
| |
Klaus Dittrich wrote:
> Nick Piggin wrote: > >> >> Your vfs_cache_pressure probably wants to be higher than 500. Make it >> 10000. > > > *No problems when using 10000. > But I think of this as workaround only. >
Yes it is just a workaround.
> I have read the "Scaling dcache with RCU" article from linuxjournal.com > to get some insight how things should work. Pretty complicated. >
Yes, prune_dcache is the only one called in response to memory pressure. Basically you wouldn't have to worry about dcache internals for your sort of machine. For example, RCU delayed freeing would not cause any problem.
The problem has probably been triggered either by a small change in the page scanner, however the underlying problem is that the slab pressure calculation isn't ideal on highmem systems, so in vfs intensive workloads (lots of files), it is easy for ZONE_NORMAL allocations to OOM without actually scanning much slab.
- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/
| |