From mboxrd@z Thu Jan 1 00:00:00 1970 From: David Teigland Date: Mon, 10 Oct 2011 15:33:46 -0400 Subject: [Cluster-devel] [Upstream patch] DLM: Convert rsb data from linked list to rb_tree In-Reply-To: <1318273207.2949.127.camel@menhir> References: <20111005200543.GD11895@redhat.com> <20111010144320.GA18764@redhat.com> <1318261861.2949.91.camel@menhir> <20111010170118.GC18764@redhat.com> <1318273207.2949.127.camel@menhir> Message-ID: <20111010193346.GD18764@redhat.com> List-Id: To: cluster-devel.redhat.com MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit On Mon, Oct 10, 2011 at 08:00:07PM +0100, Steven Whitehouse wrote: > > The fact remains that caching "as much as possible" tends to be harmful, > > and some careful limiting would be a good investment. > > > There is a limit. The point is that the limit is dynamic and depends on > memory pressure. the "as possible" part of "as much as possible". > The VM requests a reduction in the number of cached > objects when it wants to reduce the size of what we have cached. So the > larger the amount of RAM, the more inodes may be potentially cached. > > I don't agree that caching as much as possible (given the constraints > just mentioned) is bad. the "tends to" part meant that it can be good or bad, depending. > The more we cache, the more disk I/O is avoided so the better the > performance will be. You don't need to argue that more caching can be good. My point is it's not universally true in practice, as Bob's tests show.