2020-05-05 13:03:41

by Luis Henriques

[permalink] [raw]
Subject: [PATCH] ceph: demote quotarealm lookup warning to a debug message

A misconfigured cephx can easily result in having the kernel client
flooding the logs with:

ceph: Can't lookup inode 1 (err: -13)

Change his message to debug level.

Link: https://tracker.ceph.com/issues/44546
Signed-off-by: Luis Henriques <[email protected]>
---
Hi!

This patch should fix some harmless warnings when using cephx to restrict
users access to certain filesystem paths. I've added a comment to the
tracker where removing this warning could result (unlikely, IMHO!) in an
admin to miss not-so-harmless bogus configurations.

Cheers,
--
Lu?s

fs/ceph/quota.c | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/fs/ceph/quota.c b/fs/ceph/quota.c
index de56dee60540..19507e2fdb57 100644
--- a/fs/ceph/quota.c
+++ b/fs/ceph/quota.c
@@ -159,8 +159,8 @@ static struct inode *lookup_quotarealm_inode(struct ceph_mds_client *mdsc,
}

if (IS_ERR(in)) {
- pr_warn("Can't lookup inode %llx (err: %ld)\n",
- realm->ino, PTR_ERR(in));
+ dout("Can't lookup inode %llx (err: %ld)\n",
+ realm->ino, PTR_ERR(in));
qri->timeout = jiffies + msecs_to_jiffies(60 * 1000); /* XXX */
} else {
qri->timeout = 0;


2020-05-05 14:57:53

by Jeff Layton

[permalink] [raw]
Subject: Re: [PATCH] ceph: demote quotarealm lookup warning to a debug message

On Tue, 2020-05-05 at 13:59 +0100, Luis Henriques wrote:
> A misconfigured cephx can easily result in having the kernel client
> flooding the logs with:
>
> ceph: Can't lookup inode 1 (err: -13)
>
> Change his message to debug level.
>
> Link: https://tracker.ceph.com/issues/44546
> Signed-off-by: Luis Henriques <[email protected]>
> ---
> Hi!
>
> This patch should fix some harmless warnings when using cephx to restrict
> users access to certain filesystem paths. I've added a comment to the
> tracker where removing this warning could result (unlikely, IMHO!) in an
> admin to miss not-so-harmless bogus configurations.
>
> Cheers,
> --
> Luís
>
> fs/ceph/quota.c | 4 ++--
> 1 file changed, 2 insertions(+), 2 deletions(-)
>
> diff --git a/fs/ceph/quota.c b/fs/ceph/quota.c
> index de56dee60540..19507e2fdb57 100644
> --- a/fs/ceph/quota.c
> +++ b/fs/ceph/quota.c
> @@ -159,8 +159,8 @@ static struct inode *lookup_quotarealm_inode(struct ceph_mds_client *mdsc,
> }
>
> if (IS_ERR(in)) {
> - pr_warn("Can't lookup inode %llx (err: %ld)\n",
> - realm->ino, PTR_ERR(in));
> + dout("Can't lookup inode %llx (err: %ld)\n",
> + realm->ino, PTR_ERR(in));
> qri->timeout = jiffies + msecs_to_jiffies(60 * 1000); /* XXX */
> } else {
> qri->timeout = 0;
>

Merged into testing.

Thanks Luis!
--
Jeff Layton <[email protected]>

2020-05-07 13:46:30

by Jeff Layton

[permalink] [raw]
Subject: Re: [PATCH] ceph: demote quotarealm lookup warning to a debug message

On Tue, 2020-05-05 at 13:59 +0100, Luis Henriques wrote:
> A misconfigured cephx can easily result in having the kernel client
> flooding the logs with:
>
> ceph: Can't lookup inode 1 (err: -13)
>
> Change his message to debug level.
>
> Link: https://tracker.ceph.com/issues/44546
> Signed-off-by: Luis Henriques <[email protected]>
> ---
> Hi!
>
> This patch should fix some harmless warnings when using cephx to restrict
> users access to certain filesystem paths. I've added a comment to the
> tracker where removing this warning could result (unlikely, IMHO!) in an
> admin to miss not-so-harmless bogus configurations.
>
> Cheers,
> --
> Luís
>
> fs/ceph/quota.c | 4 ++--
> 1 file changed, 2 insertions(+), 2 deletions(-)
>
> diff --git a/fs/ceph/quota.c b/fs/ceph/quota.c
> index de56dee60540..19507e2fdb57 100644
> --- a/fs/ceph/quota.c
> +++ b/fs/ceph/quota.c
> @@ -159,8 +159,8 @@ static struct inode *lookup_quotarealm_inode(struct ceph_mds_client *mdsc,
> }
>
> if (IS_ERR(in)) {
> - pr_warn("Can't lookup inode %llx (err: %ld)\n",
> - realm->ino, PTR_ERR(in));
> + dout("Can't lookup inode %llx (err: %ld)\n",
> + realm->ino, PTR_ERR(in));
> qri->timeout = jiffies + msecs_to_jiffies(60 * 1000); /* XXX */
> } else {
> qri->timeout = 0;
>

Ilya,

We've had a number of reports where people get a ton of kernel log spam
when they hit this problem. I think we probably ought to mark this patch
for stable and go ahead and send it to Linus for v5.7 -- any objection?

--
Jeff Layton <[email protected]>

2020-05-07 16:27:59

by Ilya Dryomov

[permalink] [raw]
Subject: Re: [PATCH] ceph: demote quotarealm lookup warning to a debug message

On Thu, May 7, 2020 at 3:44 PM Jeff Layton <[email protected]> wrote:
>
> On Tue, 2020-05-05 at 13:59 +0100, Luis Henriques wrote:
> > A misconfigured cephx can easily result in having the kernel client
> > flooding the logs with:
> >
> > ceph: Can't lookup inode 1 (err: -13)
> >
> > Change his message to debug level.
> >
> > Link: https://tracker.ceph.com/issues/44546
> > Signed-off-by: Luis Henriques <[email protected]>
> > ---
> > Hi!
> >
> > This patch should fix some harmless warnings when using cephx to restrict
> > users access to certain filesystem paths. I've added a comment to the
> > tracker where removing this warning could result (unlikely, IMHO!) in an
> > admin to miss not-so-harmless bogus configurations.
> >
> > Cheers,
> > --
> > Luís
> >
> > fs/ceph/quota.c | 4 ++--
> > 1 file changed, 2 insertions(+), 2 deletions(-)
> >
> > diff --git a/fs/ceph/quota.c b/fs/ceph/quota.c
> > index de56dee60540..19507e2fdb57 100644
> > --- a/fs/ceph/quota.c
> > +++ b/fs/ceph/quota.c
> > @@ -159,8 +159,8 @@ static struct inode *lookup_quotarealm_inode(struct ceph_mds_client *mdsc,
> > }
> >
> > if (IS_ERR(in)) {
> > - pr_warn("Can't lookup inode %llx (err: %ld)\n",
> > - realm->ino, PTR_ERR(in));
> > + dout("Can't lookup inode %llx (err: %ld)\n",
> > + realm->ino, PTR_ERR(in));
> > qri->timeout = jiffies + msecs_to_jiffies(60 * 1000); /* XXX */
> > } else {
> > qri->timeout = 0;
> >
>
> Ilya,
>
> We've had a number of reports where people get a ton of kernel log spam
> when they hit this problem. I think we probably ought to mark this patch
> for stable and go ahead and send it to Linus for v5.7 -- any objection?

Sure, I'll queue it up.

Thanks,

Ilya