Return-Path: Received: from fieldses.org ([174.143.236.118]:33362 "EHLO fieldses.org" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751566Ab1HZUTN (ORCPT ); Fri, 26 Aug 2011 16:19:13 -0400 Date: Fri, 26 Aug 2011 16:19:10 -0400 To: Boaz Harrosh Cc: "J. Bruce Fields" , NFS list Subject: Re: [RFC] nfsd4: This is: "fix failure to end nfsd4 grace period" ++ Message-ID: <20110826201910.GB17196@fieldses.org> References: <4E45C594.1030609@panasas.com> Content-Type: text/plain; charset=us-ascii In-Reply-To: <4E45C594.1030609@panasas.com> From: "J. Bruce Fields" Sender: linux-nfs-owner@vger.kernel.org List-ID: MIME-Version: 1.0 On Fri, Aug 12, 2011 at 05:30:12PM -0700, Boaz Harrosh wrote: > Bruce hi > > This is the patch I'm currently using to be able to run. It is > your patch but the cl_firststate = 1 is promoted to before any > failures. Makes sense, thanks. > Please Also consider adding and promoting some of these prints > to DMESGs so Admins can know they have a broken setup. Like > directories do not exist or permission not properly set. OK, I agree, though a printk in every failure case seems overkill; could you live with the following? --b. commit 1f2c6beb157790cfb253e0c856e97c80e8b1bbfd Author: Boaz Harrosh Date: Fri Aug 12 17:30:12 2011 -0700 nfsd4: fix failure to end nfsd4 grace period Even if we fail to write a recovery record, we should still mark the client as having acquired its first state. Otherwise we leave 4.1 clients with indefinite ERR_GRACE returns. However, an inability to write stable storage records may cause failures of reboot recovery, and the problem should still be brought to the server administrator's attention. So, make sure the error is logged. These errors shouldn't normally be triggered on a corectly functioning server--this isn't a case where a misconfigured client could spam the logs. Signed-off-by: Boaz Harrosh Signed-off-by: J. Bruce Fields diff --git a/fs/nfsd/nfs4recover.c b/fs/nfsd/nfs4recover.c index 02eb38e..1ea241a 100644 --- a/fs/nfsd/nfs4recover.c +++ b/fs/nfsd/nfs4recover.c @@ -129,6 +129,7 @@ nfsd4_create_clid_dir(struct nfs4_client *clp) if (!rec_file || clp->cl_firststate) return 0; + clp->cl_firststate = 1; status = nfs4_save_creds(&original_cred); if (status < 0) return status; @@ -143,10 +144,8 @@ nfsd4_create_clid_dir(struct nfs4_client *clp) goto out_unlock; } status = -EEXIST; - if (dentry->d_inode) { - dprintk("NFSD: nfsd4_create_clid_dir: DIRECTORY EXISTS\n"); + if (dentry->d_inode) goto out_put; - } status = mnt_want_write(rec_file->f_path.mnt); if (status) goto out_put; @@ -156,12 +155,14 @@ out_put: dput(dentry); out_unlock: mutex_unlock(&dir->d_inode->i_mutex); - if (status == 0) { - clp->cl_firststate = 1; + if (status == 0) vfs_fsync(rec_file, 0); - } + else + printk(KERN_ERR "NFSD: failed to write recovery record" + " (err %d); please check that recovery" + " directory exists and is writeable", + status); nfs4_reset_creds(original_cred); - dprintk("NFSD: nfsd4_create_clid_dir returns %d\n", status); return status; }