[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <1302765.1628257488@warthog.procyon.org.uk>
Date: Fri, 06 Aug 2021 14:44:48 +0100
From: David Howells <dhowells@...hat.com>
To: Matthew Wilcox <willy@...radead.org>
Cc: dhowells@...hat.com, Anna Schumaker <anna.schumaker@...app.com>,
Trond Myklebust <trond.myklebust@...merspace.com>,
Jeff Layton <jlayton@...hat.com>,
Steve French <sfrench@...ba.org>,
Dominique Martinet <asmadeus@...ewreck.org>,
Mike Marshall <hubcap@...ibond.com>,
Miklos Szeredi <miklos@...redi.hu>,
Shyam Prasad N <nspmangalore@...il.com>,
Linus Torvalds <torvalds@...ux-foundation.org>,
linux-cachefs@...hat.com, linux-afs@...ts.infradead.org,
linux-nfs@...r.kernel.org, linux-cifs@...r.kernel.org,
ceph-devel@...r.kernel.org, v9fs-developer@...ts.sourceforge.net,
devel@...ts.orangefs.org, linux-mm@...ck.org,
linux-fsdevel@...r.kernel.org, linux-kernel@...r.kernel.org
Subject: Re: Canvassing for network filesystem write size vs page size
Matthew Wilcox <willy@...radead.org> wrote:
> Filesystems should not make an assumption about this ... I suspect
> the optimum page size scales with I/O bandwidth; taking PCI bandwidth
> as a reasonable proxy, it's doubled five times in twenty years.
There are a lot more factors than you make out. Local caching, content
crypto, transport crypto, cost of setting up RPC calls, compounding calls to
multiple servers.
David
Powered by blists - more mailing lists