From: Steve Dickson <SteveD@redhat.com>
To: Simo Sorce <simo@redhat.com>
Cc: "J. Bruce Fields" <bfields@fieldses.org>,
Linux NFS Mailing list <linux-nfs@vger.kernel.org>
Subject: Re: [PATCH 1/2] nfs-service: Added the starting of gssproxy
Date: Tue, 23 Sep 2014 10:58:38 -0400 [thread overview]
Message-ID: <54218A9E.8050801@RedHat.com> (raw)
In-Reply-To: <20140923085245.007533b6@willson.usersys.redhat.com>
On 09/23/2014 08:52 AM, Simo Sorce wrote:
> On Mon, 22 Sep 2014 21:19:18 -0400
> Steve Dickson <SteveD@redhat.com> wrote:
>
>>
>>
>> On 09/22/2014 08:19 PM, Simo Sorce wrote:
>>> On Mon, 22 Sep 2014 18:57:10 -0400
>>> Steve Dickson <SteveD@redhat.com> wrote:
>>>
>>>>
>>>>
>>>> On 09/22/2014 05:32 PM, Simo Sorce wrote:
>>>>> On Mon, 22 Sep 2014 17:14:05 -0400
>>>>> Steve Dickson <SteveD@redhat.com> wrote:
>>>>>
>>>>>>
>>>>>>
>>>>>> On 09/22/2014 04:44 PM, J. Bruce Fields wrote:
>>>>>>> On Mon, Sep 22, 2014 at 03:43:09PM -0400, Steve Dickson wrote:
>>>>>>>>
>>>>>>>>
>>>>>>>> On 09/22/2014 03:26 PM, Simo Sorce wrote:
>>>>>>>>> On Mon, 22 Sep 2014 15:20:07 -0400
>>>>>>>>> Steve Dickson <steved@redhat.com> wrote:
>>>>>>>>>
>>>>>>>>>> Added the gssproxy.service to both the Wants= and
>>>>>>>>>> Atfers= lines, before the rpc-svcgssd.service. There
>>>>>>>>>> are ConditionPathExists= lines in the rpc-svcgssd.service
>>>>>>>>>> unit which will stop the rpc.svcgssd daemon from
>>>>>>>>>> starting when the gssproxy daemon is already running.
>>>>>>>>>>
>>>>>>>>>> Signed-off-by: Steve Dickson <steved@redhat.com>
>>>>>>>>>> ---
>>>>>>>>>> systemd/nfs-server.service | 5 +++--
>>>>>>>>>> 1 file changed, 3 insertions(+), 2 deletions(-)
>>>>>>>>>>
>>>>>>>>>> diff --git a/systemd/nfs-server.service
>>>>>>>>>> b/systemd/nfs-server.service index 2fa7387..c740fa2 100644
>>>>>>>>>> --- a/systemd/nfs-server.service
>>>>>>>>>> +++ b/systemd/nfs-server.service
>>>>>>>>>> @@ -2,12 +2,13 @@
>>>>>>>>>> Description=NFS server and services
>>>>>>>>>> Requires= network.target proc-fs-nfsd.mount rpcbind.target
>>>>>>>>>> Requires= nfs-mountd.service
>>>>>>>>>> -Wants=rpc-statd.service nfs-idmapd.service rpc-gssd.service
>>>>>>>>>> rpc-svcgssd.service +Wants=rpc-statd.service
>>>>>>>>>> nfs-idmapd.service +Wants=rpc-gssd.service
>>>>>>>>>> Wants=rpc-statd-notify.service
>>>>>>>>>>
>>>>>>>>>> After= network.target proc-fs-nfsd.mount rpcbind.target
>>>>>>>>>> nfs-mountd.service After= nfs-idmapd.service
>>>>>>>>>> rpc-statd.service -After= rpc-gssd.service
>>>>>>>>>> rpc-svcgssd.service +After= rpc-gssd.service
>>>>>>>>>> gssproxy.service rpc-svcgssd.service Before=
>>>>>>>>>> rpc-statd-notify.service
>>>>>>>>>> Wants=nfs-config.service
>>>>>>>>>
>>>>>>>>> I think you really need to insure that the modules are loaded
>>>>>>>>> before any of the server services are started, perhaps adding
>>>>>>>>> a unit file that exec's modprobe and has "Before:
>>>>>>>>> gssproxy.service " in it ?
>>>>>>>> I really don't think its needed... From my testing it appears
>>>>>>>> gssproxy is always being started and rpc.svcgssd is not...
>>>>>>>
>>>>>>> Huh. Well rpc-svcgssd.service has var-lib-nfs-rpc_pipefs.mount
>>>>>>> as both "Requires=" and "After=", so rpc-svcgssd.service will
>>>>>>> never run without first running var-lib-nfs-rpc_pipefs.mount,
>>>>>>> which will load sunrpc. But I don't see where auth_rpcgss is
>>>>>>> getting loaded. And I don't see what ensures anything
>>>>>>> happening before gssproxy runs.
>>>>>> It happens during the mount on the client and when the server
>>>>>> is started.
>>>>>>
>>>>>>>
>>>>>>> We want to make sure your testing's not just getting lucky on
>>>>>>> the startup order.
>>>>>> The reason it working is because rpc.gssd is being started on the
>>>>>> server these days for callbacks and the After= line in
>>>>>> rpc-svcgssd.service is being executed before the
>>>>>> ConditionPathExists which cause rpc.svcgssd not to start.
>>>>>
>>>>> This guarantees ordering (to some degree) between rpc.gssd and
>>>>> rpoc.svcgssd, but says nothing about gssproxy ...
>>>> The question was how is the auth_rpcgss module being loaded. Since
>>>> both rpc-svcgssd.service and rpc-gssd.service service have
>>>> a After=var-lib-nfs-rpc_pipefs.mount and gssproxy is requiring
>>>> them, that's how auth_rpcgss is being loaded.
>>>>
>>>> If you only in enable gssproxy (not nfs-server or nfs-client) the
>>>> module still get loaded via gssproxy,service file
>>>>
>>>>>> So when gssproxy.service does it's "Before=nfs-secure.service
>>>>>> nfs-secure-server.service" line everything is loaded before
>>>>>> gssproxy start...
>>>>>>
>>>>>> I'm think gssproxy.service just needs to the put the Wants and
>>>>>> After= var-lib-nfs-rpc_pipefs.mount lines, instead of that Before
>>>>>> line..
>>>>>
>>>>> Maybe we should add "Before: gssproxy.service rpc-svcgssd.service"
>>>>> to var-lib-nfs-rpc_pipefs.mount instead (and also drop any mention
>>>>> of nfs services in gssproxy unit file so you have complete control
>>>>> of the dependencies ?
>>>> No.
>>>> The loading of sunrpc and the mounting of the file system has
>>>> nothing to do with starting up the gssd daemons.
>>>>
>>>> I would suggest gssproxy does to two things:
>>>>
>>>> 1) Add a Requires: nfs-utils to the spec file since you are
>>>> requiring services from nfs-utils
>>>
>>> No we are not requiring services from nfs-utils, nfs-utils is one of
>>> our users, you got that reversed.
>> So having "Before=nfs-secure.service nfs-secure-server.service" in
>> gssproxy.service and not having nfs-utils installed will not cause the
>> service fail during start up?
>
> It will not, but right now gssproxy.service also has:
> Requires=proc-fs-nfsd.mount
>
> I want to drop this one.
You don't want to drop this because it loads the needed kernel modules.
>
>>>> 2) Add a After=var-lib-nfs-rpc_pipefs.mount to the
>>>> gssproxy.service file since gssproxy could careless about either
>>>> rpc.gssd or rpc.svcgssd daemons. All it is looking for is the
>>>> sunrpc and auth_rpcgss kernel modules.
>>>
>>> The correct thing is to add a Before: gssproxy.service to
>>> var-lib-nfs-rpc_pipefs.mount IMO.
>> Again what happen if gssproxy is not installed? Things will
>> still come up successfully?
>
> Yes, afaik After and Before are just ordering instruction and do not
> cause any failure if the units are not present at all, or fail to start.
Ok... thanks!
steved.
>
> There are Want and Require directives for strong relationships.
>
> Simo.
>
next prev parent reply other threads:[~2014-09-23 14:58 UTC|newest]
Thread overview: 54+ messages / expand[flat|nested] mbox.gz Atom feed top
2014-09-22 19:20 [PATCH 0/2] Use the gssproxy damon for GSSAPI credentials (v3) Steve Dickson
2014-09-22 19:20 ` [PATCH 1/2] nfs-service: Added the starting of gssproxy Steve Dickson
2014-09-22 19:26 ` Simo Sorce
2014-09-22 19:43 ` Steve Dickson
2014-09-22 20:44 ` J. Bruce Fields
2014-09-22 21:14 ` Steve Dickson
2014-09-22 21:32 ` Simo Sorce
2014-09-22 22:57 ` Steve Dickson
2014-09-23 0:19 ` Simo Sorce
2014-09-23 1:19 ` Steve Dickson
2014-09-23 12:52 ` Simo Sorce
2014-09-23 14:58 ` Steve Dickson [this message]
2014-09-23 15:08 ` Simo Sorce
2014-09-23 19:29 ` J. Bruce Fields
2014-09-23 19:40 ` Simo Sorce
2014-09-23 19:51 ` J. Bruce Fields
2014-09-22 22:34 ` J. Bruce Fields
2014-09-22 23:58 ` Steve Dickson
2014-09-23 0:26 ` Simo Sorce
2014-09-23 1:55 ` J. Bruce Fields
2014-09-23 2:08 ` NeilBrown
2014-09-23 2:11 ` J. Bruce Fields
2014-09-23 19:23 ` J. Bruce Fields
2014-09-23 20:17 ` Steve Dickson
2014-09-23 20:25 ` J. Bruce Fields
2014-09-23 21:15 ` Steve Dickson
2014-09-24 15:07 ` Steve Dickson
2014-09-24 15:15 ` J. Bruce Fields
2014-09-24 15:23 ` J. Bruce Fields
2014-09-24 15:30 ` Steve Dickson
2014-09-23 12:48 ` Simo Sorce
2014-09-23 15:20 ` J. Bruce Fields
2014-09-23 16:00 ` Simo Sorce
2014-09-23 16:12 ` J. Bruce Fields
2014-09-23 16:57 ` Simo Sorce
2014-09-23 12:46 ` Simo Sorce
2014-09-23 15:06 ` Steve Dickson
2014-09-23 15:16 ` J. Bruce Fields
2014-09-23 15:52 ` Steve Dickson
2014-09-23 16:05 ` J. Bruce Fields
2014-09-23 2:01 ` NeilBrown
2014-09-22 19:40 ` J. Bruce Fields
2014-09-22 19:46 ` Simo Sorce
2014-09-22 19:53 ` Steve Dickson
2014-09-22 20:00 ` Simo Sorce
2014-09-22 20:02 ` Steve Dickson
2014-09-23 1:42 ` NeilBrown
2014-09-23 2:09 ` J. Bruce Fields
2014-09-23 2:55 ` NeilBrown
2014-09-23 12:45 ` Simo Sorce
2014-09-22 19:20 ` [PATCH 2/2] rpc.svcgssd: the build of rpc.svcgssd is off by default Steve Dickson
2014-09-22 19:43 ` J. Bruce Fields
2014-09-22 19:50 ` Steve Dickson
2014-09-22 20:21 ` J. Bruce Fields
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=54218A9E.8050801@RedHat.com \
--to=steved@redhat.com \
--cc=bfields@fieldses.org \
--cc=linux-nfs@vger.kernel.org \
--cc=simo@redhat.com \
/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
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox