2014-08-03 14:34:57

by Henrique Martins

[permalink] [raw]
Subject: nfs-utils.spec patch


The spec file for nfs-utils-1.3.0-2.1.fc20 (nfs-utils.spec)
has a bunch of bad date entries in %changelog which cause
warnings when running rpmbuild.

Patch attached, by changing the weekday to match the day
number.

-- Henrique


Attachments:
nfs-utils.spec.patch (6.28 kB)

2014-08-13 15:08:31

by Henrique Martins

[permalink] [raw]
Subject: Re: nfs-utils.spec patch

me> Patch attached, by changing the weekday to match the day
me> number.

steve> Thanks for doing this!!! I definitely appreciated it!!

You're welcome.

steve> I just did the commit on all three branches
steve> (f20,f21,master) but I didn't rebuild anything... If
steve> you need something rebuilt, just let me know...

No rebuild is needed, but maybe there should be some debate
whether to ignore exportfs errors when starting nfs as the
default, which just requires adding a couple of '-' to the
systemd service file.

A more complicated approach (would require code changes as
opposed to configuration changes) would be to not error out
if there is at least one exported mount point.

-- Henrique

2014-08-14 09:19:17

by Steve Dickson

[permalink] [raw]
Subject: Re: nfs-utils.spec patch

On 08/13/2014 11:08 AM, Henrique Martins wrote:
> No rebuild is needed, but maybe there should be some debate
> whether to ignore exportfs errors when starting nfs as the
> default, which just requires adding a couple of '-' to the
> systemd service file.
What would be the reasoning? If there is nothing exported
why come up cleanly?

>
> A more complicated approach (would require code changes as
> opposed to configuration changes) would be to not error out
> if there is at least one exported mount point.
More complexity than needed... IMHO...

steved.

2014-08-14 14:13:56

by Henrique Martins

[permalink] [raw]
Subject: Re: nfs-utils.spec patch

>> but maybe there should be some debate whether to ignore
>> exportfs errors when starting nfs as the default, which
>> just requires adding a couple of '-' to the systemd
>> service file.

> What would be the reasoning? If there is nothing exported
> why come up cleanly?

The "reasoning" is why I filed:
https://bugzilla.redhat.com/show_bug.cgi?id=1115179
and explained "it" in comment #2
https://bugzilla.redhat.com/show_bug.cgi?id=1115179#c2

I have a few machines in a laptop's /etc/exports, some only
resolve when I'm connected to the net at work, some only
resolve when I'm connected to the net at home. With the
default service file nfs always fails to start...

>> A more complicated approach (would require code changes as
>> opposed to configuration changes) would be to not error out
>> if there is at least one exported mount point.

> More complexity than needed... IMHO...

... and thus the need for the complexity to make it clean,
i.e. as long as there is a machine on the particular net I'm
connected to, that can nfs mount the laptop, nfs server
should start and not require me to do some manual editing of
/etc/exports, or write a script to take care of that.

-- Henrique

2014-08-13 13:59:27

by Steve Dickson

[permalink] [raw]
Subject: Re: nfs-utils.spec patch



On 08/03/2014 10:34 AM, Henrique Martins wrote:
>
> The spec file for nfs-utils-1.3.0-2.1.fc20 (nfs-utils.spec)
> has a bunch of bad date entries in %changelog which cause
> warnings when running rpmbuild.
>
> Patch attached, by changing the weekday to match the day
> number.

Thanks for doing this!!! I definitely appreciated it!!

I just did the commit on all three branches (f20,f21,master)
but I didn't rebuild anything... If you need something
rebuilt, just let me know...

thanks again!

steved.