sorry. Jeff Layton.
I have repush a patch and elaborate on this issue -- [PATCH] ceph: add
mount opt, always_auth
thanks for you reply.
Gregory Farnum <gfarnum@xxxxxxxxxx> 于2019年9月7日周六 上午4:45写道：
>
> On Thu, Sep 5, 2019 at 11:46 AM Jeff Layton <jlayton@xxxxxxxxxx> wrote:
> >
> > On Thu, 2019-09-05 at 08:51 -0400, simon gao wrote:
> >
> > I think this deserves a much better description than just the subject
> > line above. This change may be obvious to others but it's not clear to
> > me why this will have the effect you suggest, and what downsides it may
> > have.
> >
> > Can you flesh out the patch description and resend?
>
> Yeah this seems wrong to me in general. I don't remember exactly how
> it's implemented client-side but generally speaking we want to allow
> use of the replicated metadata on slave MDSes so that we eg spread the
> load around on subtree boundaries, and don't have to split up request
> streams across multiple MDSes when we're doing stuff like working in a
> private subdirectory but occasionally need to look at stuff on the
> parent.
> In general if the client knows which MDS is auth for a dentry/inode I
> don't think they'll send requests elsewhere without a good reason for
> it.
> -Greg
>
> >
> > Thanks,
> >
> > > ---
> > > fs/ceph/dir.c | 4 ++--
> > > fs/ceph/export.c | 8 ++++----
> > > fs/ceph/file.c | 2 +-
> > > fs/ceph/inode.c | 2 +-
> > > fs/ceph/mds_client.c | 1 +
> > > fs/ceph/super.c | 2 +-
> > > 6 files changed, 10 insertions(+), 9 deletions(-)
> > >
> > > diff --git a/fs/ceph/dir.c b/fs/ceph/dir.c
> > > index 4ca0b8f..a441b8d 100644
> > > --- a/fs/ceph/dir.c
> > > +++ b/fs/ceph/dir.c
> > > @@ -771,7 +771,7 @@ static struct dentry *ceph_lookup(struct inode *dir, struct dentry *dentry,
> > >
> > > op = ceph_snap(dir) == CEPH_SNAPDIR ?
> > > CEPH_MDS_OP_LOOKUPSNAP : CEPH_MDS_OP_LOOKUP;
> > > - req = ceph_mdsc_create_request(mdsc, op, USE_ANY_MDS);
> > > + req = ceph_mdsc_create_request(mdsc, op, USE_AUTH_MDS);
> > > if (IS_ERR(req))
> > > return ERR_CAST(req);
> > > req->r_dentry = dget(dentry);
> > > @@ -1600,7 +1600,7 @@ static int ceph_d_revalidate(struct dentry *dentry, unsigned int flags)
> > >
> > > op = ceph_snap(dir) == CEPH_SNAPDIR ?
> > > CEPH_MDS_OP_LOOKUPSNAP : CEPH_MDS_OP_LOOKUP;
> > > - req = ceph_mdsc_create_request(mdsc, op, USE_ANY_MDS);
> > > + req = ceph_mdsc_create_request(mdsc, op, USE_AUTH_MDS);
> > > if (!IS_ERR(req)) {
> > > req->r_dentry = dget(dentry);
> > > req->r_num_caps = 2;
> > > diff --git a/fs/ceph/export.c b/fs/ceph/export.c
> > > index 15ff1b0..a7d5174 100644
> > > --- a/fs/ceph/export.c
> > > +++ b/fs/ceph/export.c
> > > @@ -135,7 +135,7 @@ static struct inode *__lookup_inode(struct super_block *sb, u64 ino)
> > > int mask;
> > >
> > > req = ceph_mdsc_create_request(mdsc, CEPH_MDS_OP_LOOKUPINO,
> > > - USE_ANY_MDS);
> > > + USE_AUTH_MDS);
> > > if (IS_ERR(req))
> > > return ERR_CAST(req);
> > >
> > > @@ -210,7 +210,7 @@ static struct dentry *__snapfh_to_dentry(struct super_block *sb,
> > > return d_obtain_alias(inode);
> > >
> > > req = ceph_mdsc_create_request(mdsc, CEPH_MDS_OP_LOOKUPINO,
> > > - USE_ANY_MDS);
> > > + USE_AUTH_MDS);
> > > if (IS_ERR(req))
> > > return ERR_CAST(req);
> > >
> > > @@ -294,7 +294,7 @@ static struct dentry *__get_parent(struct super_block *sb,
> > > int err;
> > >
> > > req = ceph_mdsc_create_request(mdsc, CEPH_MDS_OP_LOOKUPPARENT,
> > > - USE_ANY_MDS);
> > > + USE_AUTH_MDS);
> > > if (IS_ERR(req))
> > > return ERR_CAST(req);
> > >
> > > @@ -509,7 +509,7 @@ static int ceph_get_name(struct dentry *parent, char *name,
> > >
> > > mdsc = ceph_inode_to_client(inode)->mdsc;
> > > req = ceph_mdsc_create_request(mdsc, CEPH_MDS_OP_LOOKUPNAME,
> > > - USE_ANY_MDS);
> > > + USE_AUTH_MDS);
> > > if (IS_ERR(req))
> > > return PTR_ERR(req);
> > >
> > > diff --git a/fs/ceph/file.c b/fs/ceph/file.c
> > > index 685a03c..79533f2 100644
> > > --- a/fs/ceph/file.c
> > > +++ b/fs/ceph/file.c
> > > @@ -182,7 +182,7 @@ static void put_bvecs(struct bio_vec *bvecs, int num_bvecs, bool should_dirty)
> > > struct ceph_fs_client *fsc = ceph_sb_to_client(sb);
> > > struct ceph_mds_client *mdsc = fsc->mdsc;
> > > struct ceph_mds_request *req;
> > > - int want_auth = USE_ANY_MDS;
> > > + int want_auth = USE_AUTH_MDS;
> > > int op = (flags & O_CREAT) ? CEPH_MDS_OP_CREATE : CEPH_MDS_OP_OPEN;
> > >
> > > if (flags & (O_WRONLY|O_RDWR|O_CREAT|O_TRUNC))
> > > diff --git a/fs/ceph/inode.c b/fs/ceph/inode.c
> > > index 18500ede..6c67548 100644
> > > --- a/fs/ceph/inode.c
> > > +++ b/fs/ceph/inode.c
> > > @@ -2247,7 +2247,7 @@ int __ceph_do_getattr(struct inode *inode, struct page *locked_page,
> > > if (!force && ceph_caps_issued_mask(ceph_inode(inode), mask, 1))
> > > return 0;
> > >
> > > - mode = (mask & CEPH_STAT_RSTAT) ? USE_AUTH_MDS : USE_ANY_MDS;
> > > + mode = USE_AUTH_MDS;
> > > req = ceph_mdsc_create_request(mdsc, CEPH_MDS_OP_GETATTR, mode);
> > > if (IS_ERR(req))
> > > return PTR_ERR(req);
> > > diff --git a/fs/ceph/mds_client.c b/fs/ceph/mds_client.c
> > > index 920e9f0..acfb969 100644
> > > --- a/fs/ceph/mds_client.c
> > > +++ b/fs/ceph/mds_client.c
> > > @@ -867,6 +867,7 @@ static struct inode *get_nonsnap_parent(struct dentry *dentry)
> > > return inode;
> > > }
> > >
> > > +static struct inode *get_parent()
> > > /*
> > > * Choose mds to send request to next. If there is a hint set in the
> > > * request (e.g., due to a prior forward hint from the mds), use that.
> > > diff --git a/fs/ceph/super.c b/fs/ceph/super.c
> > > index ab4868c..517e605 100644
> > > --- a/fs/ceph/super.c
> > > +++ b/fs/ceph/super.c
> > > @@ -867,7 +867,7 @@ static struct dentry *open_root_dentry(struct ceph_fs_client *fsc,
> > >
> > > /* open dir */
> > > dout("open_root_inode opening '%s'\n", path);
> > > - req = ceph_mdsc_create_request(mdsc, CEPH_MDS_OP_GETATTR, USE_ANY_MDS);
> > > + req = ceph_mdsc_create_request(mdsc, CEPH_MDS_OP_GETATTR, USE_AUTH_MDS);
> > > if (IS_ERR(req))
> > > return ERR_CAST(req);
> > > req->r_path1 = kstrdup(path, GFP_NOFS);
> >
> >
> > --
> > Jeff Layton <jlayton@xxxxxxxxxx>
> >
