From mboxrd@z Thu Jan 1 00:00:00 1970 MIME-Version: 1.0 In-Reply-To: <150029e986ff7e10958868a983ff9794@quanstro.net> References: <20090421161033.170FD5B24@mail.bitblocks.com> <150029e986ff7e10958868a983ff9794@quanstro.net> Date: Tue, 21 Apr 2009 18:23:28 +0100 Message-ID: From: roger peppe To: Fans of the OS Plan 9 from Bell Labs <9fans@9fans.net> Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: quoted-printable Subject: Re: [9fans] Plan9 - the next 20 years Topicbox-Message-UUID: ee86e216-ead4-11e9-9d60-3106f5b1d025 2009/4/21 erik quanstrom : > bundling is equivalent to running the original sequence on > the remote machine and shipping only the result back. =C2=A0some > rtt latency is eliminated but i think things will still be largely > in-order because walks will act like fences. =C2=A0i think the lots- > of-small-files case will still suffer. =C2=A0maybe i'm not quite followin= g > along. i agree that the lots-of-small-files case will still suffer (mainly because the non-hierarchical mount table means we can't know what's mounted below a particular node without asking the server). but this still gives the opportunity to considerably speed up many common actions (e.g. {walk, open, read, close}) without adding too much (i think) complexity of the protocol. also, as david leimbach points out, this case is still amenable to the "send several requests concurrently" approach.