namespace documentation.

Eric W. Biederman ebiederm at xmission.com
Fri Dec 21 22:58:56 UTC 2012


Rob Landley <rob at landley.net> writes:

> On 12/21/2012 11:51:03 AM, Eric W. Biederman wrote:
>> Oleg Nesterov <oleg at redhat.com> writes:
>> 
>> > Eric. I understand that it is too late to discuss this. And yes, I  
>> simply
>> > do not understand the problem space, I never used containers.
>> >
>> > But, stupid question. Let's ignore the pid_ns-specific oddities.
>> >
>> > 1. Ignoring setns(), why do we need /proc/pid/ns/ ?
>> >
>> > 2. Why setns() requires /proc/pid/ns/ ? IOW, why it can't be
>> >
>> > 	sys_setns(pid_t pid, int clone_flags)
>> > 	{
>> > 		truct task_struct *tsk = find_task_by_vpid(pid);
>> > 		struct nsproxy *target = get_nsproxy(tsk->nsproxy);
>> >
>> > 		new_nsproxy = create_new_namespaces(...);
>> >
>> > 		if (clone_flags & CLONE_NEWNS)
>> > 			mntns_install(...);
>> > 		if (clone_flags & CLONE_NEWIPC)
>> > 			ipcns_install(...);
>> > 		...
>> > 	}
>> >
>> > I feel I missed something trivial, but what?
>> 
>> It is a question of naming.
>> 
>> The problem I set out to solve when all of this was introduced was how
>> to name namespaces without introducing yet another namespace.
>> 
>> The solution to the naming problem that I finally found was to  
>> introduce
>> something I could mount.
>
> Where might I find documentation on this? I'm aware of  
> Documentation/namespaces but it's only got one file in it (about  
> conflicts between namespace types). I'm aware of  
> http://lxc.sourceforge.net/index.php/about/kernel-namespaces/ and  
> http://lxc.sourceforge.net/man/ but that's mixed in with the  
> implementation details of a particular userspace tool, and tends to lag  
> the kernel significantly. (Those man pages were last updated in 2010,  
> which if I recall was the last time I poked them about it.)

I'm not certain what you are asking about.

The man pages that I endeavour to keep reasonably current are.

man 5 proc
man 2 setns
man 2 unshare
man 2 clone

You won't get a design discussion but you will get a description of how
the existing pieces work.  Of course now that I look it appears my
patches have not merged yet.  But that is reasonable since my recent
changes did not merge until a few days ago.

There is also iproute2 it's man pages and source.

There is the kernel source.

There are the occassional lwn articles.

I believe there should be a reasonable amount of email in the mailing
list archives when talking about the design descision, and when I
introduced setns.

Eric


More information about the Containers mailing list