Re: 2.6.6-rc{1,2} bad VM/NFS interaction in case of dirty page writeback

[Shantanu Goel]

--- Andrew Morton <akpm@xxxxxxxx> wrote:
> Trond Myklebust <trond.myklebust@xxxxxxxxxx> wrote:
> >
> > 
> > Err... Anything that currently ends up calling
> writepage() and returning
> > WRITEPAGE_ACTIVATE. This is a problem that you
> believe you are seeing
> > the effects of, right? 
> 
> I didn't report the problem - Shantanu is reporting
> problems where all the
> NFS pages are stuck on the active list and the VM
> has nothing to reclaim on
> the inactive list.
> 

Yup, what happens is the NFS writepage() returns
WRITEPAGE_ACTIVATE causing the scanner to place the
page at the head of the active list.  Now the page
won't be reclaimed until the scanner has run through
the entire active list.  I do not see such behaviour
with ext3 for instance.



	
		
__________________________________
Do you Yahoo!?
Win a $20,000 Career Makeover at Yahoo! HotJobs  
http://hotjobs.sweepstakes.yahoo.com/careermakeover 
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/