ceph-devel.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Jeff Layton <jlayton@kernel.org>
To: Luis Henriques <lhenriques@suse.de>
Cc: ceph-devel@vger.kernel.org, idryomov@gmail.com,
	Sage Weil <sage@redhat.com>
Subject: Re: [PATCH] ceph: add a new vxattr to return auth mds for an inode
Date: Tue, 27 Jul 2021 11:53:18 -0400	[thread overview]
Message-ID: <e7c531436cc7f66c9599ee7426f28af80b2a1fb5.camel@kernel.org> (raw)
In-Reply-To: <YQAqHRX3Y52QSsFf@suse.de>

On Tue, 2021-07-27 at 16:45 +0100, Luis Henriques wrote:
> On Tue, Jul 27, 2021 at 11:17:39AM -0400, Jeff Layton wrote:
> > On Tue, 2021-07-27 at 15:46 +0100, Luis Henriques wrote:
> > > On Tue, Jul 27, 2021 at 07:35:09AM -0400, Jeff Layton wrote:
> > > > Add a new vxattr that shows what MDS is authoritative for an inode (if
> > > > we happen to have auth caps). If we don't have an auth cap for the inode
> > > > then just return -1.
> > > > 
> > > > URL: https://tracker.ceph.com/issues/1276
> > > > Signed-off-by: Jeff Layton <jlayton@kernel.org>
> > > > ---
> > > >  fs/ceph/xattr.c | 16 ++++++++++++++++
> > > >  1 file changed, 16 insertions(+)
> > > > 
> > > > diff --git a/fs/ceph/xattr.c b/fs/ceph/xattr.c
> > > > index 1242db8d3444..70664a19b8dc 100644
> > > > --- a/fs/ceph/xattr.c
> > > > +++ b/fs/ceph/xattr.c
> > > > @@ -340,6 +340,15 @@ static ssize_t ceph_vxattrcb_caps(struct ceph_inode_info *ci, char *val,
> > > >  			      ceph_cap_string(issued), issued);
> > > >  }
> > > >  
> > > > +static ssize_t ceph_vxattrcb_auth_mds(struct ceph_inode_info *ci,
> > > > +				       char *val, size_t size)
> > > > +{
> > > > +	/* return -1 if we don't have auth caps (and thus can't tell) */
> > > > +	if (!ci->i_auth_cap)
> > > > +		return ceph_fmt_xattr(val, size, "-1");
> > > 
> > > I don't really have an opinion on this as I don't have a usecase for this
> > > xattr (other than debug, of course).  But I just checked a similar function
> > > ceph_vxattrcb_layout_pool_namespace() and, if there's no value for ns for an
> > > inode, it just returns 0.
> > > 
> > > Anyway, just my 5c, as I'm OK with returning a '-1' string too.
> > > 
> > 
> > 
> > TBH, I don't have much of a use-case for this either, but it was
> > requested by Sage long ago. That said, I figure it might be useful in
> > some cases, particularly when troubleshooting pinning issues.
> > 
> > Pool numbering is a bit different, as I think pool 0 is not valid,
> > whereas we index mds's starting with 0. I'm fine with a different
> > convention here, but I considered it as a safe way to say "I don't know"
> > in this situation.
> 
> Right, but what I meant by returning 0 was to really do:
> 
> 	if (!ci->i_auth_cap)
> 		return 0;
> 
> and not to return the string "0".  Anyway, as I said, I don't have an
> opinion on this and returning the "-1" string for the MDS index sounds
> good too.
> 
> Cheers,
> --
> Luís
> 

Oh yeah, ok. We could do that I guess. I don't care much either way, but
that might be cleaner. Let me think on it a bit.

I also think we'll need to hold the ci->i_ceph_lock around the
i_auth_cap access. I'll need to send a v2 patch for that anyway.

> > > > +	return ceph_fmt_xattr(val, size, "%d", ci->i_auth_cap->session->s_mds);
> > > > +}
> > > > +
> > > >  #define CEPH_XATTR_NAME(_type, _name)	XATTR_CEPH_PREFIX #_type "." #_name
> > > >  #define CEPH_XATTR_NAME2(_type, _name, _name2)	\
> > > >  	XATTR_CEPH_PREFIX #_type "." #_name "." #_name2
> > > > @@ -473,6 +482,13 @@ static struct ceph_vxattr ceph_common_vxattrs[] = {
> > > >  		.exists_cb = NULL,
> > > >  		.flags = VXATTR_FLAG_READONLY,
> > > >  	},
> > > > +	{
> > > > +		.name = "ceph.auth_mds",
> > > > +		.name_size = sizeof("ceph.auth_mds"),
> > > > +		.getxattr_cb = ceph_vxattrcb_auth_mds,
> > > > +		.exists_cb = NULL,
> > > > +		.flags = VXATTR_FLAG_READONLY,
> > > > +	},
> > > >  	{ .name = NULL, 0 }	/* Required table terminator */
> > > >  };
> > > >  
> > > > -- 
> > > > 2.31.1
> > > > 
> > 
> > -- 
> > Jeff Layton <jlayton@kernel.org>
> > 

-- 
Jeff Layton <jlayton@kernel.org>


  reply	other threads:[~2021-07-27 15:53 UTC|newest]

Thread overview: 7+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2021-07-27 11:35 Jeff Layton
2021-07-27 14:46 ` Luis Henriques
2021-07-27 15:17   ` Jeff Layton
2021-07-27 15:45     ` Luis Henriques
2021-07-27 15:53       ` Jeff Layton [this message]
2021-07-27 18:42 ` [PATCH v2] " Jeff Layton
2021-07-28 10:28   ` Luis Henriques

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=e7c531436cc7f66c9599ee7426f28af80b2a1fb5.camel@kernel.org \
    --to=jlayton@kernel.org \
    --cc=ceph-devel@vger.kernel.org \
    --cc=idryomov@gmail.com \
    --cc=lhenriques@suse.de \
    --cc=sage@redhat.com \
    --subject='Re: [PATCH] ceph: add a new vxattr to return auth mds for an inode' \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).