Xen project Mailing List

On Aug 14, 2016 5:06 AM, <xen-users-request@xxxxxxxxxxxxx> wrote:

Send Xen-users mailing list submissions to
xen-users@xxxxxxxxxxxxx

To subscribe or unsubscribe via the World Wide Web, visit
https://lists.xen.org/cgi-bin/mailman/listinfo/xen-users
or, via email, send a message with subject or body 'help' to
xen-users-request@xxxxxxxxx.org

You can reach the person managing the list at
xen-users-owner@xxxxxxxxxxxxx

When replying, please edit your Subject line so it is more specific
than "Re: Contents of Xen-users digest..."

Today's Topics:

1. xen networking problem (=?ISO-8859-1?B?ZmFuZ3R1bzkw?=)
2. Re: Openindiana (was Mac OSX How-to) (jim burns)
3. Re: xen networking problem (Sarah Newman)
4. Re: xen networking problem (Big Strong)
5. Re: xen networking problem (Manfred Haertel, DB3HM)
6. Re: xen networking problem (Big Strong)

----------------------------------------------------------------------

Message: 1
Date: Sat, 13 Aug 2016 20:57:04 +0800
From: "=?ISO-8859-1?B?ZmFuZ3R1bzkw?=" <fangtuo90@xxxxxxxxx>
To: "=?ISO-8859-1?B?eGVuLXVzZXJz?=" <xen-users@xxxxxxxxxxxxx>
Subject: [Xen-users] xen networking problem
Message-ID: <tencent_73E199A9555C1EA72DB4125C@qq.com>
Content-Type: text/plain; charset="iso-8859-1"

I'm using openvswitch as the bridge. And for the VM dom1, two interfaces are created, vif1.0 and vif1.0-emu. Both have bytes received and transmited, but not equal. Which one represents the traffic of dom1? It is needed to be classified to adjust the route of the VM based on it's ports, or dump traffics for analysis.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.xen.org/archives/html/xen-users/attachments/20160813/06d656ff/attachment.html>

------------------------------

Message: 2
Date: Sat, 13 Aug 2016 10:33:57 -0400
From: jim burns <jim_burn@xxxxxxxxxxxxx>
To: PÁSZTOR György <pasztor@linux.gyakg.u-szeged.hu>
Cc: xen-users@xxxxxxxxxxxxx
Subject: Re: [Xen-users] Openindiana (was Mac OSX How-to)
Message-ID: <3967172.X0AdaZLVXU@insp3847>
Content-Type: text/plain; charset="iso-8859-1"

On Saturday, 13 August 2016, 12:16:22 EDT, PÁSZTOR György wrote:
> Hi,
>
> "jim burns" <jim_burn@xxxxxxxxxxxxx> írta 2016-08-12 22:07-kor:
> > Yeah, that's pretty much Boris's method. When I execute the cfg for the
> > installer, I get: (if reading this at home, make sure your email client is
> > reading in html mode)
> >
> > xl create -c oi
> > Parsing config from oi
> > libxl: warning: libxl_bootloader.c:416:bootloader_disk_attached_cb:
> > bootloader='/usr/bin/pygrub' is deprecated; use bootloader='pygrub'
> > instead
> > v4.7.0 chgset ''
>
> Which xen version did you tried?

As it says above, 4.7.0.

> Maybe something broke in xen during times

That's what I think too. Boris' examples would have been around late xen 3.x,
early 4.x, maybe 4.1.

> As I mentioned, I tried this in
> ~2013-2014. What I did not mentioned: I tried that on a stable or oldstable
> debian as dom0 os. In those times debian 6.0 was the oldstable, 7.0 was the
> stable. So the newset out of the box xen version was 4.1. And after a short
> try we gave up on use of the xl toolstack: With xl, if you said reboot in a
> guest system, then it did not rebooted, just become dead somewhere.
> I filled a Debian bug report in their bts. They said, they will not fix it,
> since supporting "xl" toolstack is "experimental".
> I could not win over their stubbornness :(
> https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=727100
>
> Btw.: Did you tried openindiana only? I am not sure, maybe I just dreamed
> that, but I slighlty remember, if I'd read that on omnios-discuss, that
> there is available omnios AMI images in AWS. So, if it is possible, than
> maybe omnios's kernel / other things are fixed to use in a xen environment.
>
> Cheers,
> Gyu

Not familiar with Omnios. How similar to OS is it?

Thanx.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.xen.org/archives/html/xen-users/attachments/20160813/a28417a2/attachment.html>

------------------------------

Message: 3
Date: Sat, 13 Aug 2016 12:48:56 -0700
From: Sarah Newman <srn@xxxxxxxxx>
To: fangtuo90 <fangtuo90@xxxxxxxxx>
Cc: xen-users <xen-users@xxxxxxxxxxxxx>
Subject: Re: [Xen-users] xen networking problem
Message-ID: <abb9d1b2-a2cf-1806-7811-20bf098da577@xxxxxxxxx>
Content-Type: text/plain; charset=utf-8

On 08/13/2016 05:57 AM, fangtuo90 wrote:
> I'm using openvswitch as the bridge. And for the VM dom1, two interfaces are created, vif1.0 and vif1.0-emu. Both have bytes received and transmited, but not equal. Which one represents the traffic of dom1? It is needed to be classified to adjust the route of the VM based on it's ports, or dump traffics for analysis.

Both.

https://wiki.xen.org/wiki/Xen_Networking#Emulated_Network_Devices

------------------------------

Message: 4
Date: Sun, 14 Aug 2016 09:17:00 +0800
From: Big Strong <fangtuo90@xxxxxxxxx>
To: Sarah Newman <srn@xxxxxxxxx>
Cc: xen-users <xen-users@xxxxxxxxxxxxx>
Subject: Re: [Xen-users] xen networking problem
Message-ID:
<CAFnE1f2A9qZ9F+Fh1+7in_rudpKYdL9se9jL6DOsYPV9arL6qA@mail.gmail.com>
Content-Type: text/plain; charset="utf-8"

I'm still confused as the picture
http://xenbits.xen.org/docs/unstable-staging/figs/network-bridge.png shows
no vif4.0-emu, and as I said, the traffic of vif and vif-emu is different,
which is verified by wireshark. There are many interfaces shown in dom0 by
ifconfig, Is the relationship of these interfaces as follows?
host-eth0 -- xenbr0 -- vif -- vif-emu -- vm-eth0

2016-08-14 3:48 GMT+08:00 Sarah Newman <srn@xxxxxxxxx>:

> On 08/13/2016 05:57 AM, fangtuo90 wrote:
> > I'm using openvswitch as the bridge. And for the VM dom1, two interfaces
> are created, vif1.0 and vif1.0-emu. Both have bytes received and
> transmited, but not equal. Which one represents the traffic of dom1? It is
> needed to be classified to adjust the route of the VM based on it's ports,
> or dump traffics for analysis.
>
> Both.
>
> https://wiki.xen.org/wiki/Xen_Networking#Emulated_Network_Devices
>
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.xen.org/archives/html/xen-users/attachments/20160814/ed82a211/attachment.html>

------------------------------

Message: 5
Date: Sun, 14 Aug 2016 05:54:18 +0200
From: "Manfred Haertel, DB3HM" <Manfred.Haertel@xxxxxxxxxxxx>
To: xen-users <xen-users@xxxxxxxxxxxxx>
Subject: Re: [Xen-users] xen networking problem
Message-ID: <57AFEB6A.9010700@xxxxxxxxxxxx>
Content-Type: text/plain; charset=UTF-8; format=flowed

Sarah Newman schrieb:
> On 08/13/2016 05:57 AM, fangtuo90 wrote:
>> I'm using openvswitch as the bridge. And for the VM dom1, two interfaces are created, vif1.0 and vif1.0-emu. Both have bytes received and transmited, but not equal. Which one represents the traffic of dom1? It is needed to be classified to adjust the route of the VM based on it's ports, or dump traffics for analysis.
>
> Both.
>
> https://wiki.xen.org/wiki/Xen_Networking#Emulated_Network_Devices

The VM *could* use both vif devices, but typically it decides for one of
them and most often this is the paravirtualized one. E.g. the Linux
kernel, whenn running in a Xen VM, does so.

However, since the Xen DOM0 does not know for which interface the VM
would decide, it typically connects *both* Devices to the virtual
bridge. This does not harm, if one of them is not used.

Because the virtual bridge has to flood broadcast packets to all its
ports, typically you can see quite a few *transmitted* packets also on
the unused device. On the other hand, the number of *received* packets
should be 0, and it is on my Xen DOM0 (just checked).

--
Manfred Härtel, DB3HM mailto:Manfred.Haertel@rz-online.de
http://rz-home.de/mhaertel

------------------------------

Message: 6
Date: Sun, 14 Aug 2016 15:17:19 +0800
From: Big Strong <fangtuo90@xxxxxxxxx>
To: "Manfred Haertel, DB3HM" <Manfred.Haertel@xxxxxxxxxxxx>
Cc: xen-users <xen-users@xxxxxxxxxxxxx>
Subject: Re: [Xen-users] xen networking problem
Message-ID:
<CAFnE1f2We0b_vdCXAg2-Ew7FekNOXYncpj04EadqRAHPk2Z_xg@xxxxxxxxxxxxxx>
Content-Type: text/plain; charset="utf-8"

Thanks very much for the clear explanation.

2016-08-14 11:54 GMT+08:00 Manfred Haertel, DB3HM <
Manfred.Haertel@xxxxxxxxxxxx>:

> Sarah Newman schrieb:
>
> On 08/13/2016 05:57 AM, fangtuo90 wrote:
>>
>>> I'm using openvswitch as the bridge. And for the VM dom1, two interfaces
>>> are created, vif1.0 and vif1.0-emu. Both have bytes received and
>>> transmited, but not equal. Which one represents the traffic of dom1? It is
>>> needed to be classified to adjust the route of the VM based on it's ports,
>>> or dump traffics for analysis.
>>>
>>
>> Both.
>>
>> https://wiki.xen.org/wiki/Xen_Networking#Emulated_Network_Devices
>>
>
> The VM *could* use both vif devices, but typically it decides for one of
> them and most often this is the paravirtualized one. E.g. the Linux kernel,
> whenn running in a Xen VM, does so.
>
> However, since the Xen DOM0 does not know for which interface the VM would
> decide, it typically connects *both* Devices to the virtual bridge. This
> does not harm, if one of them is not used.
>
> Because the virtual bridge has to flood broadcast packets to all its
> ports, typically you can see quite a few *transmitted* packets also on the
> unused device. On the other hand, the number of *received* packets should
> be 0, and it is on my Xen DOM0 (just checked).
>
> --
> Manfred Härtel, DB3HM mailto:Manfred.Haertel@rz-online.de
> http://rz-home.de/mhaertel
>
> _______________________________________________
> Xen-users mailing list
> Xen-users@xxxxxxxxxxxxx
> https://lists.xen.org/xen-users
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.xen.org/archives/html/xen-users/attachments/20160814/070f4577/attachment.html>

------------------------------

Subject: Digest Footer

_______________________________________________
Xen-users mailing list
Xen-users@xxxxxxxxxxxxx
https://lists.xen.org/xen-users

------------------------------

End of Xen-users Digest, Vol 138, Issue 11
******************************************

Re: [Xen-users] Xen-users Digest, Vol 138, Issue 11