From: Herbert Poetzl <herbert@13thfloor.at>
To: "Eric W. Biederman" <ebiederm@xmission.com>
Cc: Daniel Lezcano <dlezcano@fr.ibm.com>,
Andrey Savochkin <saw@swsoft.com>,
linux-kernel@vger.kernel.org, netdev@vger.kernel.org,
serue@us.ibm.com, haveblue@us.ibm.com, clg@fr.ibm.com,
Andrew Morton <akpm@osdl.org>,
dev@sw.ru, devel@openvz.org, sam@vilain.net,
viro@ftp.linux.org.uk, Alexey Kuznetsov <alexey@sw.ru>
Subject: Re: [patch 2/6] [Network namespace] Network device sharing by view
Date: Tue, 27 Jun 2006 18:02:42 +0200 [thread overview]
Message-ID: <20060627160241.GB28984@MAIL.13thfloor.at> (raw)
In-Reply-To: <m1sllqn7cb.fsf@ebiederm.dsl.xmission.com>
On Tue, Jun 27, 2006 at 05:52:52AM -0600, Eric W. Biederman wrote:
> Daniel Lezcano <dlezcano@fr.ibm.com> writes:
>
> >>>>My point is that if you make namespace tagging at routing time,
> >>>>and your packets are being routed only once, you lose the ability
> >>>>to have separate routing tables in each namespace.
> >>>
> >>>Right. What is the advantage of having separate the routing tables ?
> >> Routing is everything. For example, I want namespaces to have their
> >> private tunnel devices. It means that namespaces should be allowed
> >> have private routes of local type, private default routes, and so
> >> on...
> >>
> >
> > Ok, we are talking about the same things. We do it only in a different way:
> >
> > * separate routing table :
> > namespace
> > |
> > \--- route_tables
> > |
> > \---routes
> >
> > * tagged routing table :
> > route_tables
> > |
> > \---routes
> > |
> > \---namespace
>
> There is a third possibility, that falls in between these two if local
> communication is really the bottle neck.
>
> We have the dst cache for caching routes and cache multiple
> transformations that happen on a packet.
>
> With a little extra knowledge it is possible to have the separate
> routing tables but have special logic that recognizes the local
> tunnel device that connects namespaces and have it look into the next
> namespaces routes, and build up a complete stack of dst entries of
> where the packet needs to go.
>
> I keep forgetting about that possibility. But as long as everything is
> done at the routing layer that should work.
>
> > I use the second method, because I think it is more effecient and
> > reduce the overhead. But the isolation is minimalist and only aims
> > to avoid the application using ressources outside of the container
> > (aka namespace) without taking care of the system. For example, I
> > didn't take care of network devices, because as far as see I can't
> > imagine an administrator wanting to change the network device name
> > while there are hundred of containers running. Concerning tunnel
> > devices for example, they should be created inside the container.
>
> Inside the containers I want all network devices named eth0!
huh? even if there are two of them? also tun?
I think you meant, you want to be able to have eth0 in
_more_ than one guest where eth0 in a guest can also
be/use/relate to eth1 on the host, right?
> > I think, private network ressources method is more elegant
> > and involves more network ressources, but there is probably a
> > significant overhead and some difficulties to have __lightweight__
> > container (aka application container), make nfs working well,
> > etc... I did some tests with tbench and the loopback with the
> > private namespace and there is roughly an overhead of 4 % without
> > the isolation since with the tagging method there is 1 % with the
> > isolation.
>
> The overhead went down?
yes, this might actually happen, because the guest
has only to look at a certain subset of entries
but this needs a lot more testing, especially with
a lot of guests
> > The network namespace aims the isolation for now, but the container
> > based on the namespaces will probably need checkpoint/restart and
> > migration ability. The migration is needed not only for servers but
> > for HPC jobs too.
>
> Yes.
>
> > So I don't know what level of isolation/virtualization is really
> > needed by users, what should be acceptable (strong isolation and
> > overhead / weak isolation and efficiency). I don't know if people
> > wanting strong isolation will not prefer Xen (cleary with much more
> > overhead than your patches ;) )
well, Xen claims something below 2% IIRC, and would
be clearly the better choice if you want strict
separation with the complete functionality, especially
with hardware support
> We need a clean abstraction that optimizes well.
>
> However local communication between containers is not what we
> should benchmark. That can always be improved later. So long as
> the performance is reasonable. What needs to be benchmarked is the
> overhead of namespaces when connected to physical networking devices
> and on their own local loopback, and comparing that to a kernel
> without namespace support.
well, for me (obviously advocating the lightweight case)
it seems improtant that the following conditions are met:
- loopback traffic inside a guest is insignificantly
slower than on a normal system
- loopback traffic on the host is insignificantly
slower than on a normal system
- inter guest traffic is faster than on-wire traffic,
and should be withing a small tolerance of the
loopback case (as it really isn't different)
- network (on-wire) traffic should be as fast as without
the namespace (i.e. within 1% or so, better not really
measurable)
- all this should be true in a setup with a significant
number of guests, when only one guest is active, but
all other guests are ready/configured
- all this should scale well with a few hundred guests
> If we don't hurt that core case we have an implementation we can
> merge. There are a lot of optimization opportunities for local
> communications and we can do that after we have a correct and accepted
> implementation. Anything else is optimizing too soon, and will
> just be muddying the waters.
what I fear is that once something is in, the kernel will
just become slower (as it already did in some areas) and
nobody will care/be-able to fix that later on ...
best,
Herbert
> Eric
next prev parent reply other threads:[~2006-06-27 16:02 UTC|newest]
Thread overview: 113+ messages / expand[flat|nested] mbox.gz Atom feed top
2006-06-09 21:02 [RFC] [patch 0/6] [Network namespace] introduction dlezcano
2006-06-09 21:02 ` [RFC] [patch 1/6] [Network namespace] Network namespace structure dlezcano
2006-06-09 21:02 ` [RFC] [patch 2/6] [Network namespace] Network device sharing by view dlezcano
2006-06-11 10:18 ` Andrew Morton
2006-06-18 18:53 ` Al Viro
2006-06-26 9:47 ` Andrey Savochkin
2006-06-26 13:02 ` Herbert Poetzl
2006-06-26 14:05 ` Eric W. Biederman
2006-06-26 14:08 ` Andrey Savochkin
2006-06-26 18:28 ` Herbert Poetzl
2006-06-26 18:59 ` Eric W. Biederman
2006-06-26 14:56 ` Daniel Lezcano
2006-06-26 15:21 ` Eric W. Biederman
2006-06-26 15:27 ` Andrey Savochkin
2006-06-26 15:49 ` Daniel Lezcano
2006-06-26 16:40 ` Eric W. Biederman
2006-06-26 18:36 ` Herbert Poetzl
2006-06-26 19:35 ` Eric W. Biederman
2006-06-26 20:02 ` Herbert Poetzl
2006-06-26 20:37 ` Eric W. Biederman
2006-06-26 21:26 ` Herbert Poetzl
2006-06-26 21:59 ` Ben Greear
2006-06-26 22:11 ` Eric W. Biederman
2006-06-27 9:09 ` Andrey Savochkin
2006-06-27 15:48 ` Herbert Poetzl
2006-06-27 16:19 ` Andrey Savochkin
2006-06-27 16:40 ` Eric W. Biederman
2006-06-26 22:13 ` Ben Greear
2006-06-26 22:54 ` Herbert Poetzl
2006-06-26 23:08 ` Ben Greear
2006-06-27 16:07 ` Ben Greear
2006-06-27 22:48 ` Herbert Poetzl
2006-06-27 9:11 ` Andrey Savochkin
2006-06-27 9:34 ` Daniel Lezcano
2006-06-27 9:38 ` Andrey Savochkin
2006-06-27 11:21 ` Daniel Lezcano
2006-06-27 11:52 ` Eric W. Biederman
2006-06-27 16:02 ` Herbert Poetzl [this message]
2006-06-27 16:47 ` Eric W. Biederman
2006-06-27 17:19 ` Ben Greear
2006-06-27 22:52 ` Herbert Poetzl
2006-06-27 23:12 ` Dave Hansen
2006-06-27 23:42 ` Alexey Kuznetsov
2006-06-28 3:38 ` Eric W. Biederman
2006-06-28 13:36 ` Herbert Poetzl
2006-06-28 13:53 ` jamal
2006-06-28 14:19 ` Andrey Savochkin
2006-06-28 16:17 ` jamal
2006-06-28 16:58 ` Andrey Savochkin
2006-06-28 17:17 ` Eric W. Biederman
2006-06-28 17:04 ` Herbert Poetzl
2006-06-28 14:39 ` Eric W. Biederman
2006-06-30 1:41 ` Sam Vilain
2006-06-29 21:07 ` Sam Vilain
2006-06-29 22:14 ` strict isolation of net interfaces Cedric Le Goater
2006-06-30 2:39 ` Serge E. Hallyn
2006-06-30 2:49 ` Sam Vilain
2006-07-03 14:53 ` Andrey Savochkin
2006-07-04 3:00 ` Sam Vilain
2006-07-04 12:29 ` Daniel Lezcano
2006-07-04 13:13 ` Sam Vilain
2006-07-04 13:19 ` Daniel Lezcano
2006-06-30 8:56 ` Cedric Le Goater
2006-07-03 13:36 ` Herbert Poetzl
2006-06-30 12:23 ` Daniel Lezcano
2006-06-30 14:20 ` Eric W. Biederman
2006-06-30 15:22 ` Daniel Lezcano
2006-06-30 17:58 ` Eric W. Biederman
2006-06-30 16:14 ` Serge E. Hallyn
2006-06-30 17:41 ` Eric W. Biederman
2006-06-30 18:09 ` Eric W. Biederman
2006-06-30 0:15 ` [patch 2/6] [Network namespace] Network device sharing by view jamal
2006-06-30 3:35 ` Herbert Poetzl
2006-06-30 7:45 ` Andrey Savochkin
2006-06-30 13:50 ` jamal
2006-06-30 15:01 ` Andrey Savochkin
2006-06-30 18:22 ` Eric W. Biederman
2006-06-30 21:51 ` jamal
2006-07-01 0:50 ` Eric W. Biederman
2006-06-28 14:21 ` Eric W. Biederman
2006-06-28 14:51 ` Eric W. Biederman
2006-06-27 16:49 ` Alexey Kuznetsov
2006-06-27 11:55 ` Andrey Savochkin
2006-06-27 9:54 ` Kirill Korotaev
2006-06-27 16:09 ` Herbert Poetzl
2006-06-27 16:29 ` Eric W. Biederman
2006-06-27 23:07 ` Herbert Poetzl
2006-06-28 4:07 ` Eric W. Biederman
2006-06-28 6:31 ` Sam Vilain
2006-06-28 14:15 ` Herbert Poetzl
2006-06-28 15:36 ` Eric W. Biederman
2006-06-28 17:18 ` Herbert Poetzl
2006-06-28 10:14 ` Cedric Le Goater
2006-06-28 14:11 ` Herbert Poetzl
2006-06-28 16:10 ` Eric W. Biederman
2006-07-06 9:45 ` Routing tables (Re: [patch 2/6] [Network namespace] Network device sharing by view) Kari Hurtta
2006-06-09 21:02 ` [RFC] [patch 3/6] [Network namespace] Network devices isolation dlezcano
2006-06-18 18:57 ` Al Viro
2006-06-09 21:02 ` [RFC] [patch 4/6] [Network namespace] Network inet " dlezcano
2006-06-09 21:02 ` [RFC] [patch 5/6] [Network namespace] ipv4 isolation dlezcano
2006-06-10 0:23 ` James Morris
2006-06-10 0:27 ` Rick Jones
2006-06-10 0:47 ` James Morris
2006-06-09 21:02 ` [RFC] [patch 6/6] [Network namespace] Network namespace debugfs dlezcano
2006-06-10 7:16 ` [RFC] [patch 0/6] [Network namespace] introduction Kari Hurtta
2006-06-16 4:23 ` Eric W. Biederman
2006-06-16 9:06 ` Daniel Lezcano
2006-06-16 9:22 ` Eric W. Biederman
2006-06-18 18:47 ` Al Viro
2006-06-20 21:21 ` Daniel Lezcano
2006-06-20 21:25 ` Al Viro
2006-06-20 22:45 ` Daniel Lezcano
2006-06-26 23:38 ` Patrick McHardy
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=20060627160241.GB28984@MAIL.13thfloor.at \
--to=herbert@13thfloor.at \
--cc=akpm@osdl.org \
--cc=alexey@sw.ru \
--cc=clg@fr.ibm.com \
--cc=dev@sw.ru \
--cc=devel@openvz.org \
--cc=dlezcano@fr.ibm.com \
--cc=ebiederm@xmission.com \
--cc=haveblue@us.ibm.com \
--cc=linux-kernel@vger.kernel.org \
--cc=netdev@vger.kernel.org \
--cc=sam@vilain.net \
--cc=saw@swsoft.com \
--cc=serue@us.ibm.com \
--cc=viro@ftp.linux.org.uk \
/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
all inboxes | Powered by JetHome®