Hello,
we use Linux NFS clients with a Linux NFS server in an configuration
where NFS mounts are done on client boot _and_ on user login in a
session; umounts are done on users logout from the session.
We see occasionally several different problems which all may have
the same root cause:
- When a client accesses a file which was accessed before
from the same client in a previous session the server
prevents access to the file until a timeout happens.
The timeout has a duration of about 1-3 minutes.
In this case the "blocked" file can not even be deleted
on the server.
--> What causes this timeout? I found nothing in the
server code which has such a timeout How can I debug what
the server is waiting for or why he is blocking access
to the file?
- Sometimes client processes hang in the middle of a session
on some file. After a timeout the file is accessible again.
The timeout can take 1 upto several minutes. The file is
also blocked on the server, it cannot be accessed.
I think all theses problemes are caused by something like
dangling locks or another invalid state on the server.
The clients show no network error like dropped packets
or something like this.
--> How can I debug such hangs?
We use Linux NFS server and client from vanilla kernel 4.4.31
with sec=sys.
Can anyone help? Does "a bell ring"?
Thank you and best regards
-Ulrich
--
| Ulrich Gemkow
| University of Stuttgart
| Institute of Communication Networks and Computer Engineering (IKR)
On Thu, Nov 17, 2016 at 09:32:47PM +0100, Ulrich Gemkow wrote:
> Hello,
>
> we use Linux NFS clients with a Linux NFS server in an configuration
> where NFS mounts are done on client boot _and_ on user login in a
> session; umounts are done on users logout from the session.
>
> We see occasionally several different problems which all may have
> the same root cause:
>
> - When a client accesses a file which was accessed before
> from the same client in a previous session the server
> prevents access to the file until a timeout happens.
>
> The timeout has a duration of about 1-3 minutes.
> In this case the "blocked" file can not even be deleted
> on the server.
>
> --> What causes this timeout? I found nothing in the
> server code which has such a timeout How can I debug what
> the server is waiting for or why he is blocking access
> to the file?
>
> - Sometimes client processes hang in the middle of a session
> on some file. After a timeout the file is accessible again.
> The timeout can take 1 upto several minutes. The file is
> also blocked on the server, it cannot be accessed.
>
> I think all theses problemes are caused by something like
> dangling locks or another invalid state on the server.
>
> The clients show no network error like dropped packets
> or something like this.
>
> --> How can I debug such hangs?
>
> We use Linux NFS server and client from vanilla kernel 4.4.31
> with sec=sys.
>
> Can anyone help? Does "a bell ring"?
The lease period is 90 seconds by default, and there are several cases
where you can end up waiting for a lease period.
For example, if the client held some delegations that it didn't return
on unmount, and then it denied knowledge of them when the server tried
to recall them, then the server would have to wait a lease period to
forcibly remove them. But, the client should be returning delegations
on unmount, so I don't see how this happens.
For locks and opens and other state, again the client should be
returning them on unmount. And anyway the server isn't going to
forcibly remove those ever, unless the entire client goes away
completely, e.g. in a client crash or network partition.
So, I don't know. Are you sure there aren't client crashes or network
problems?
Also I'd personally try to arrange things so you, say, just mount /home/
on boot instead of automounting /home/bfields when bfields logs in.
But, I don't know your situation.
--b.
Hello Bruce,
thanks...
On Thursday 17 November 2016, J. Bruce Fields wrote:
> On Thu, Nov 17, 2016 at 09:32:47PM +0100, Ulrich Gemkow wrote:
> > Hello,
> >
> > we use Linux NFS clients with a Linux NFS server in an configuration
> > where NFS mounts are done on client boot _and_ on user login in a
> > session; umounts are done on users logout from the session.
> >
> > We see occasionally several different problems which all may have
> > the same root cause:
> >
> > - When a client accesses a file which was accessed before
> > from the same client in a previous session the server
> > prevents access to the file until a timeout happens.
> >
> > The timeout has a duration of about 1-3 minutes.
> > In this case the "blocked" file can not even be deleted
> > on the server.
> >
> > --> What causes this timeout? I found nothing in the
> > server code which has such a timeout How can I debug what
> > the server is waiting for or why he is blocking access
> > to the file?
> >
> > - Sometimes client processes hang in the middle of a session
> > on some file. After a timeout the file is accessible again.
> > The timeout can take 1 upto several minutes. The file is
> > also blocked on the server, it cannot be accessed.
> >
> > I think all theses problemes are caused by something like
> > dangling locks or another invalid state on the server.
> >
> > The clients show no network error like dropped packets
> > or something like this.
> >
> > --> How can I debug such hangs?
> >
> > We use Linux NFS server and client from vanilla kernel 4.4.31
> > with sec=sys.
> >
> > Can anyone help? Does "a bell ring"?
>
> The lease period is 90 seconds by default, and there are several cases
> where you can end up waiting for a lease period.
I found the 90sec lease time period but the timeout is sometimes
much longer than 90 sec, often up to 3minutes or longer. Is there
something which may cause these longer delays (I played with the
90sec constant and it did not help :-)
> For example, if the client held some delegations that it didn't return
> on unmount, and then it denied knowledge of them when the server tried
> to recall them, then the server would have to wait a lease period to
> forcibly remove them. But, the client should be returning delegations
> on unmount, so I don't see how this happens.
>
> For locks and opens and other state, again the client should be
> returning them on unmount. And anyway the server isn't going to
> forcibly remove those ever, unless the entire client goes away
> completely, e.g. in a client crash or network partition.
>
> So, I don't know. Are you sure there aren't client crashes or network
> problems?
It happens that clients crash but IMHO the server should notice this
by dropped connections. We have no network problems in these cases.
> Also I'd personally try to arrange things so you, say, just mount /home/
> on boot instead of automounting /home/bfields when bfields logs in.
> But, I don't know your situation.
Sure, we can do this. But we are in an unsecure environment and it
gives additional (required) security to use more specific mounts
(we make the export on the server when the user has authenticated
with our own daemon).
What I really miss is an option to disable locks in NFSv4. Maybe
you can point me to the right place in the source..?
Any more ideas?
Thanks again and best regards
-Ulrich
>
> --b.
>
--
|-----------------------------------------------------------------------
| Ulrich Gemkow
|-----
| Universit?t Stuttgart
| Institut f?r Kommunikationsnetze und Rechnersysteme (IKR)
|-----
| University of Stuttgart
| Institute of Communication Networks and Computer Engineering (IKR)
|-----
| Pfaffenwaldring 47, D 70569 Stuttgart, Germany
| mailto:[email protected] http://www.ikr.uni-stuttgart.de
|-----------------------------------------------------------------------
On Thu, Nov 17, 2016 at 10:34:20PM +0100, Ulrich Gemkow wrote:
> Hello Bruce,
>
> thanks...
>
> On Thursday 17 November 2016, J. Bruce Fields wrote:
> > On Thu, Nov 17, 2016 at 09:32:47PM +0100, Ulrich Gemkow wrote:
> > > Hello,
> > >
> > > we use Linux NFS clients with a Linux NFS server in an configuration
> > > where NFS mounts are done on client boot _and_ on user login in a
> > > session; umounts are done on users logout from the session.
> > >
> > > We see occasionally several different problems which all may have
> > > the same root cause:
> > >
> > > - When a client accesses a file which was accessed before
> > > from the same client in a previous session the server
> > > prevents access to the file until a timeout happens.
> > >
> > > The timeout has a duration of about 1-3 minutes.
> > > In this case the "blocked" file can not even be deleted
> > > on the server.
> > >
> > > --> What causes this timeout? I found nothing in the
> > > server code which has such a timeout How can I debug what
> > > the server is waiting for or why he is blocking access
> > > to the file?
> > >
> > > - Sometimes client processes hang in the middle of a session
> > > on some file. After a timeout the file is accessible again.
> > > The timeout can take 1 upto several minutes. The file is
> > > also blocked on the server, it cannot be accessed.
> > >
> > > I think all theses problemes are caused by something like
> > > dangling locks or another invalid state on the server.
> > >
> > > The clients show no network error like dropped packets
> > > or something like this.
> > >
> > > --> How can I debug such hangs?
> > >
> > > We use Linux NFS server and client from vanilla kernel 4.4.31
> > > with sec=sys.
> > >
> > > Can anyone help? Does "a bell ring"?
> >
> > The lease period is 90 seconds by default, and there are several cases
> > where you can end up waiting for a lease period.
>
> I found the 90sec lease time period but the timeout is sometimes
> much longer than 90 sec, often up to 3minutes or longer. Is there
> something which may cause these longer delays (I played with the
> 90sec constant and it did not help :-)
A delegation is the only thing that I can think of that would prevent a
file from being deleted on the server (by that you mean, not even a "rm
blockfiled" run from a terminal on the server works?) Delegations
should definitely be forcibly revoked after the lease period passes.
Note that you need to reboot (well, restart the nfs server) after
changing the lease period, or the change will not take effect.
>
> > For example, if the client held some delegations that it didn't return
> > on unmount, and then it denied knowledge of them when the server tried
> > to recall them, then the server would have to wait a lease period to
> > forcibly remove them. But, the client should be returning delegations
> > on unmount, so I don't see how this happens.
> >
> > For locks and opens and other state, again the client should be
> > returning them on unmount. And anyway the server isn't going to
> > forcibly remove those ever, unless the entire client goes away
> > completely, e.g. in a client crash or network partition.
> >
> > So, I don't know. Are you sure there aren't client crashes or network
> > problems?
>
> It happens that clients crash
I'm not sure what you mean there--do you mean clients are involved in
all of these cases, or some of them?
> but IMHO the server should notice this by dropped connections. We have
> no network problems in these cases.
By design, an NFS server won't drop locks on loss a TCP connection.
They'll be dropped either:
- after a full lease period passes without the server hearing
anything from the client, or
- if the client crashes and reboots; in this case the client
should inform the server that it just rebooted and that all
its old locks can be discarded.
>
> > Also I'd personally try to arrange things so you, say, just mount /home/
> > on boot instead of automounting /home/bfields when bfields logs in.
> > But, I don't know your situation.
>
> Sure, we can do this. But we are in an unsecure environment and it
> gives additional (required) security to use more specific mounts
> (we make the export on the server when the user has authenticated
> with our own daemon).
>
> What I really miss is an option to disable locks in NFSv4. Maybe
> you can point me to the right place in the source..?
Delegations can be turned off, by running this on the server before
starting it:
echo 0 >/proc/sys/fs/leases-enable
There's no way to turn off file locks.
--b.
Hello Bruce,
On Friday 18 November 2016, J. Bruce Fields wrote:
> On Thu, Nov 17, 2016 at 10:34:20PM +0100, Ulrich Gemkow wrote:
> > Hello Bruce,
> >
> > thanks...
> >
> > On Thursday 17 November 2016, J. Bruce Fields wrote:
> > > On Thu, Nov 17, 2016 at 09:32:47PM +0100, Ulrich Gemkow wrote:
> > > > Hello,
> > > >
> > > > we use Linux NFS clients with a Linux NFS server in an configuration
> > > > where NFS mounts are done on client boot _and_ on user login in a
> > > > session; umounts are done on users logout from the session.
> > > >
> > > > We see occasionally several different problems which all may have
> > > > the same root cause:
> > > >
> > > > - When a client accesses a file which was accessed before
> > > > from the same client in a previous session the server
> > > > prevents access to the file until a timeout happens.
> > > >
> > > > The timeout has a duration of about 1-3 minutes.
> > > > In this case the "blocked" file can not even be deleted
> > > > on the server.
> > > >
> > > > --> What causes this timeout? I found nothing in the
> > > > server code which has such a timeout How can I debug what
> > > > the server is waiting for or why he is blocking access
> > > > to the file?
> > > >
> > > > - Sometimes client processes hang in the middle of a session
> > > > on some file. After a timeout the file is accessible again.
> > > > The timeout can take 1 upto several minutes. The file is
> > > > also blocked on the server, it cannot be accessed.
> > > >
> > > > I think all theses problemes are caused by something like
> > > > dangling locks or another invalid state on the server.
> > > >
> > > > The clients show no network error like dropped packets
> > > > or something like this.
> > > >
> > > > --> How can I debug such hangs?
> > > >
> > > > We use Linux NFS server and client from vanilla kernel 4.4.31
> > > > with sec=sys.
> > > >
> > > > Can anyone help? Does "a bell ring"?
> > >
> > > The lease period is 90 seconds by default, and there are several cases
> > > where you can end up waiting for a lease period.
> >
> > I found the 90sec lease time period but the timeout is sometimes
> > much longer than 90 sec, often up to 3minutes or longer. Is there
> > something which may cause these longer delays (I played with the
> > 90sec constant and it did not help :-)
>
> A delegation is the only thing that I can think of that would prevent a
> file from being deleted on the server (by that you mean, not even a "rm
> blockfiled" run from a terminal on the server works?) Delegations
> should definitely be forcibly revoked after the lease period passes.
> Note that you need to reboot (well, restart the nfs server) after
> changing the lease period, or the change will not take effect.
Thanks for this hint, I will disable delegations. But - the timeout
is for sure longer than 90 seconds in many cases. Can the reason be
a bad interaction between dropped tcp-connections (which may require
some time to be noticed) and the nfs server state(s)?
> > > For example, if the client held some delegations that it didn't return
> > > on unmount, and then it denied knowledge of them when the server tried
> > > to recall them, then the server would have to wait a lease period to
> > > forcibly remove them. But, the client should be returning delegations
> > > on unmount, so I don't see how this happens.
> > >
> > > For locks and opens and other state, again the client should be
> > > returning them on unmount. And anyway the server isn't going to
> > > forcibly remove those ever, unless the entire client goes away
> > > completely, e.g. in a client crash or network partition.
> > >
> > > So, I don't know. Are you sure there aren't client crashes or network
> > > problems?
> >
> > It happens that clients crash
>
> I'm not sure what you mean there--do you mean clients are involved in
> all of these cases, or some of them?
Cause for the client reboots are impatient users which switch power
off-and-on when a hang happens. So the crashes (reboots) are not
directly related but the hangs happen often after such unwanted
reboots.
> > but IMHO the server should notice this by dropped connections. We have
> > no network problems in these cases.
>
> By design, an NFS server won't drop locks on loss a TCP connection.
> They'll be dropped either:
>
> - after a full lease period passes without the server hearing
> anything from the client, or
> - if the client crashes and reboots; in this case the client
> should inform the server that it just rebooted and that all
> its old locks can be discarded.
>
> >
> > > Also I'd personally try to arrange things so you, say, just mount /home/
> > > on boot instead of automounting /home/bfields when bfields logs in.
> > > But, I don't know your situation.
> >
> > Sure, we can do this. But we are in an unsecure environment and it
> > gives additional (required) security to use more specific mounts
> > (we make the export on the server when the user has authenticated
> > with our own daemon).
> >
> > What I really miss is an option to disable locks in NFSv4. Maybe
> > you can point me to the right place in the source..?
>
> Delegations can be turned off, by running this on the server before
> starting it:
>
> echo 0 >/proc/sys/fs/leases-enable
>
> There's no way to turn off file locks.
>
> --b.
>
Thanks again and best regards!
-Ulrich
--
|-----------------------------------------------------------------------
| Ulrich Gemkow
| University of Stuttgart
| Institute of Communication Networks and Computer Engineering (IKR)
|-----------------------------------------------------------------------
On Fri, Nov 18, 2016 at 07:55:50PM +0100, Ulrich Gemkow wrote:
> On Friday 18 November 2016, J. Bruce Fields wrote:
> > On Thu, Nov 17, 2016 at 10:34:20PM +0100, Ulrich Gemkow wrote:
> > > I found the 90sec lease time period but the timeout is sometimes
> > > much longer than 90 sec, often up to 3minutes or longer. Is there
> > > something which may cause these longer delays (I played with the
> > > 90sec constant and it did not help :-)
> >
> > A delegation is the only thing that I can think of that would prevent a
> > file from being deleted on the server (by that you mean, not even a "rm
> > blockfiled" run from a terminal on the server works?) Delegations
> > should definitely be forcibly revoked after the lease period passes.
> > Note that you need to reboot (well, restart the nfs server) after
> > changing the lease period, or the change will not take effect.
>
> Thanks for this hint, I will disable delegations. But - the timeout
> is for sure longer than 90 seconds in many cases. Can the reason be
> a bad interaction between dropped tcp-connections (which may require
> some time to be noticed) and the nfs server state(s)?
If the problem is a delegation, then what happens is essentially:
- you try to modify (or rename, or remove) the delated file.
- the server sets a timer for the lease time (90s by default).
- at the same time, the server notifies the client that it
should return the delegation.
- if the timer expires then the server gives up and forcibly
removes the delegation, allowing your original operation to
proceed.
So tcp connections and stuff are details, what matters to the server is
how much time has elapsed since you attempted an operation that
conflicts with the delegation. If that's significantly more than the
lease period, then something's wrong. So if you have a case where
that's reliably too long, that would be interesting.
> > > > For example, if the client held some delegations that it didn't return
> > > > on unmount, and then it denied knowledge of them when the server tried
> > > > to recall them, then the server would have to wait a lease period to
> > > > forcibly remove them. But, the client should be returning delegations
> > > > on unmount, so I don't see how this happens.
> > > >
> > > > For locks and opens and other state, again the client should be
> > > > returning them on unmount. And anyway the server isn't going to
> > > > forcibly remove those ever, unless the entire client goes away
> > > > completely, e.g. in a client crash or network partition.
> > > >
> > > > So, I don't know. Are you sure there aren't client crashes or network
> > > > problems?
> > >
> > > It happens that clients crash
> >
> > I'm not sure what you mean there--do you mean clients are involved in
> > all of these cases, or some of them?
>
> Cause for the client reboots are impatient users which switch power
> off-and-on when a hang happens. So the crashes (reboots) are not
> directly related but the hangs happen often after such unwanted
> reboots.
Hm. So their stale state should be cleared out either 90 seconds after
the client turned off, or as soon as the client comes back up and
remounts, whichever comes first. If that's not happening, again that
sounds like a potentially interesting bug.
--b.
Hallo Bruce,
On Friday 18 November 2016, J. Bruce Fields wrote:
> On Fri, Nov 18, 2016 at 07:55:50PM +0100, Ulrich Gemkow wrote:
> > On Friday 18 November 2016, J. Bruce Fields wrote:
> > > On Thu, Nov 17, 2016 at 10:34:20PM +0100, Ulrich Gemkow wrote:
> > > > I found the 90sec lease time period but the timeout is sometimes
> > > > much longer than 90 sec, often up to 3minutes or longer. Is there
> > > > something which may cause these longer delays (I played with the
> > > > 90sec constant and it did not help :-)
> > >
> > > A delegation is the only thing that I can think of that would prevent a
> > > file from being deleted on the server (by that you mean, not even a "rm
> > > blockfiled" run from a terminal on the server works?) Delegations
> > > should definitely be forcibly revoked after the lease period passes.
> > > Note that you need to reboot (well, restart the nfs server) after
> > > changing the lease period, or the change will not take effect.
> >
> > Thanks for this hint, I will disable delegations. But - the timeout
> > is for sure longer than 90 seconds in many cases. Can the reason be
> > a bad interaction between dropped tcp-connections (which may require
> > some time to be noticed) and the nfs server state(s)?
>
> If the problem is a delegation, then what happens is essentially:
>
> - you try to modify (or rename, or remove) the delated file.
> - the server sets a timer for the lease time (90s by default).
> - at the same time, the server notifies the client that it
> should return the delegation.
> - if the timer expires then the server gives up and forcibly
> removes the delegation, allowing your original operation to
> proceed.
>
> So tcp connections and stuff are details, what matters to the server is
> how much time has elapsed since you attempted an operation that
> conflicts with the delegation. If that's significantly more than the
> lease period, then something's wrong. So if you have a case where
> that's reliably too long, that would be interesting.
ok, thanks for the explanation, this helps a lot to understand
the correlations. I will disable delegations and see whether
something changed. I will come back then to this.
> > > > > For example, if the client held some delegations that it didn't return
> > > > > on unmount, and then it denied knowledge of them when the server tried
> > > > > to recall them, then the server would have to wait a lease period to
> > > > > forcibly remove them. But, the client should be returning delegations
> > > > > on unmount, so I don't see how this happens.
> > > > >
> > > > > For locks and opens and other state, again the client should be
> > > > > returning them on unmount. And anyway the server isn't going to
> > > > > forcibly remove those ever, unless the entire client goes away
> > > > > completely, e.g. in a client crash or network partition.
> > > > >
> > > > > So, I don't know. Are you sure there aren't client crashes or network
> > > > > problems?
> > > >
> > > > It happens that clients crash
> > >
> > > I'm not sure what you mean there--do you mean clients are involved in
> > > all of these cases, or some of them?
> >
> > Cause for the client reboots are impatient users which switch power
> > off-and-on when a hang happens. So the crashes (reboots) are not
> > directly related but the hangs happen often after such unwanted
> > reboots.
>
> Hm. So their stale state should be cleared out either 90 seconds after
> the client turned off, or as soon as the client comes back up and
> remounts, whichever comes first. If that's not happening, again that
> sounds like a potentially interesting bug.
Currently I only see the delay without having a chance to further
debug this because I cannnot see the "inner state" of client and/or
server. I will try to use the debug messages to see more.
Thanks again and best regards
-Ulrich
>
> --b.
> --
> To unsubscribe from this list: send the line "unsubscribe linux-nfs" in
> the body of a message to [email protected]
> More majordomo info at http://vger.kernel.org/majordomo-info.html
--
| Ulrich Gemkow
| University of Stuttgart
| Institute of Communication Networks and Computer Engineering (IKR)
|-----------------------------------------------------------------------