Moving to autofs.

Before we go further can you try autofs revision 38 please.

You can get revision 38 at:
http://kojipkgs.fedoraproject.org/packages/autofs/5.0.5/38.fc15

or wait until it is pushed to the testing repo.

(In reply to comment #2)
> Before we go further can you try autofs revision 38 please.
> 
> You can get revision 38 at:
> http://kojipkgs.fedoraproject.org/packages/autofs/5.0.5/38.fc15
> 
> or wait until it is pushed to the testing repo.

Just upgraded to the following version:
autofs.i686        1:5.0.5-38.fc15         @updates-testing

and the issue persists.

(In reply to comment #3)
I have seen the same problem with 1:5.0.5-38.fc15 of autofs and the solution of changing 

  # Required-Start: $network $ypbind
  # Required-Stop: $network $ypbind

to

  # Required-Start: $network ypbind
  # Required-Stop: $network ypbind

seems to have solved it for me too - reading the Fedora documentation I did not find any mentioning of what the $ signs mean in this context but judging from 

http://refspecs.freestandards.org/LSB_3.1.0/LSB-Core-generic/LSB-Core-generic/facilname.html

Nick's solution seems to be the correct fix for the problem?

I would suggest adding a section about the 'reserved special facilities' in the fedora docs since it seems to have confused others as well: 

ypbind currently uses 

# Required-Start: $local_fs $remote_fs $network $rpcbind

and as far as I understand rpcbind is also not a 'special facility' ?

I think this is a duplicate of Bug 709637 (?)

Is there a change that this bug gets fixed in F16? It's very annoying and according to Comment #3 the fix would be simple.

The fix described here is still needed in F16.

I've checked with "systemctl dot", and with the "$ypbind" form autofs has a dependency of "ypbind.target", which is plainly wrong.  With s/$ypbind/ypbind/, the dependency is correctly on "ypbind.service".

This is a simple fix, but autofs over NIS is completely non-functional without it; can we get it applied soon, please?

(The fix is necessary but not sufficient, unfortunately; we also need a fix for ypbind, to wait for the service to be functional before marking it ready, which I'll report separately.  So "systemctl dot" is a better way to be sure that this particular fix is working.)

(In reply to comment #7)
> The fix described here is still needed in F16.
> 
> I've checked with "systemctl dot", and with the "$ypbind" form autofs has a
> dependency of "ypbind.target", which is plainly wrong.  With s/$ypbind/ypbind/,
> the dependency is correctly on "ypbind.service".

Ahh .. a common sense reasoned observation at last.

> 
> This is a simple fix, but autofs over NIS is completely non-functional without
> it; can we get it applied soon, please?

Yes, I'll try to get it done today.

Ian

Can we give this scratch build a try please.

It can be found at:
http://people.redhat.com/~ikent/autofs-5.0.5-39.fc15

(In reply to comment #8)
> (In reply to comment #7)
> > I've checked with "systemctl dot", and with the "$ypbind" form autofs has a
> > dependency of "ypbind.target", which is plainly wrong.  With s/$ypbind/ypbind/,
> > the dependency is correctly on "ypbind.service".
> 
> Ahh .. a common sense reasoned observation at last.

"At last"?  I was sure that I mentioned that s/$ypbind/ypbind/ corrects the issue in the original report...

(In reply to comment #9)
> Can we give this scratch build a try please.
> 
> It can be found at:
> http://people.redhat.com/~ikent/autofs-5.0.5-39.fc15

Unsurprisingly, since this version includes the aforementioned fix, it
works.

(In reply to comment #10)
> (In reply to comment #8)
> > (In reply to comment #7)
> > > I've checked with "systemctl dot", and with the "$ypbind" form autofs has a
> > > dependency of "ypbind.target", which is plainly wrong.  With s/$ypbind/ypbind/,
> > > the dependency is correctly on "ypbind.service".
> > 
> > Ahh .. a common sense reasoned observation at last.
> 
> "At last"?  I was sure that I mentioned that s/$ypbind/ypbind/ corrects the
> issue in the original report...

Yes, but what you may not be aware of is that this has gone on
for quite a long time and I have made changes along the way based
on people saying "this appears to fix the problem" and it actually
didn't.

I didn't know what the root cause of this was but Stephen offered
the observation of what was happening within systemd which made
sense.

It was (and possibly still will be) frustrating for you but it
was frustrating form me well before.

Anyway, that's for checking the change, I'll commit the change
and request a push and update the other distros.

Ian

(In reply to comment #11)
> 
> Anyway, that's for checking the change, I'll commit the change
> and request a push and update the other distros.

That should be thanks, not that's.

> 
> Ian

autofs-5.0.5-39.fc15 has been submitted as an update for Fedora 15.
https://admin.fedoraproject.org/updates/autofs-5.0.5-39.fc15

autofs-5.0.6-4.fc16 has been submitted as an update for Fedora 16.
https://admin.fedoraproject.org/updates/autofs-5.0.6-4.fc16

Package autofs-5.0.5-39.fc15:
* should fix your issue,
* was pushed to the Fedora 15 testing repository,
* should be available at your local mirror within two days.
Update it with:
# su -c 'yum update --enablerepo=updates-testing autofs-5.0.5-39.fc15'
as soon as you are able to.
Please go to the following url:
https://admin.fedoraproject.org/updates/FEDORA-2011-16695/autofs-5.0.5-39.fc15
then log in and leave karma (feedback).

Thanks, f16 packages from koji fix this part of the problem for me.  I'll pursue the other problems separately (ypbind and rpcbind both need initscripts/systemd fixes too.)

autofs-5.0.6-4.fc16 has been pushed to the Fedora 16 stable repository.  If problems still persist, please make note of it in this bug report.

autofs-5.0.5-39.fc15 has been pushed to the Fedora 15 stable repository.  If problems still persist, please make note of it in this bug report.