Return-Path: Received: from daytona.panasas.com ([67.152.220.89]:28257 "EHLO daytona.panasas.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1754934Ab1EKPrc (ORCPT ); Wed, 11 May 2011 11:47:32 -0400 Message-ID: <4DCAC586.2040205@panasas.com> Date: Wed, 11 May 2011 13:21:10 -0400 From: Benny Halevy To: chao liu CC: bwickman@citi.umich.edu, linux-nfs@vger.kernel.org Subject: Re: some questions about NFSv4.1_Directory_Delegation References: In-Reply-To: Content-Type: text/plain; charset=ISO-8859-1 Sender: linux-nfs-owner@vger.kernel.org List-ID: MIME-Version: 1.0 On 2011-05-11 05:30, chao liu wrote: > *hi everyone,* > * **Benny and **bwickman**,* > * i have read the > article https://wiki.linux-nfs.org/wiki/index.php/CITI_Experience_with_Directory_Delegations > carefully,* > * and i checked the latest source code in git hub, i can't > find the code about directory delegation , so i got two questions to ask,. *ny > * > * > * To Benny, thank you very much, do you think that what are > the most difficulties or problems to implement Directory_Delegation in > pnfs?* Directory delegations are not related to pNFS but rather to NFSv4.1. I wasn't involved in their experimental implementation in CITI so I don't know exactly what were their challenges. I can only guess that the harder parts were dealing with the name cache rather than implementing the related NFSv4.1 code. As far as I know, the benefits demonstrated so far are not good enough to justify the added complexity. Benny > * > * > * To the author who wrote the article(the author may know > bwickman), also thanks very much, you said that you have implemented * > * Directory_Delegation prototype, how can i find the sourcecode? * > * * > * > * > > >