[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Xen-users] Xen networking disconnect



Hi,

Thanks, I’ll try; my host is a CentOS 5/Xen3 pre installed.

 

Julio.

 


De: Stephan Seitz [mailto:s.seitz@xxxxxxxxxxx]
Enviado el: Lunes, 30 de Julio de 2012 12:52 p.m.
Para: Julio Solórzano
CC: xen-users@xxxxxxxxxxxxx
Asunto: Re: [Xen-users] Xen networking disconnect

 

Hi,

as there's Marks "networking experiments" mail attached to your's, I'm quite not sure if you're
using the legacy (as of your CentOS 5/Xen3 setup) way, or maybe trying the experimental one.
I guess, the attached mail is unrelated, otherwise I'ld have probably noticed openvswitch.

Install ethtool inside your domU and try

ethtool -K eth0 tx off # assuming eth0 is your domU's interface to xenbr0

if that helps, you could add that line in your domU's interfaces (below the respective iface statement):
    post-up  ethtool -K eth0 tx off

cheers,

Stephan



Am Montag, den 30.07.2012, 10:50 -0500 schrieb Julio Solórzano:

 
Hello.
I need your help, DomU xen network diconnects randomly, how do you
monitoring this?
When I connect to vnc server in the hosts, DomU is running, then I open the
DomU, after that guest network responds. 
DomU Debian 6.03 over Centos 5.7 and Xen 3.03.
My network configuration:
eth0      Link encap:Ethernet  HWaddr 68:B5:99:78:AA:6D
          inet addr:192.168.5.230  Bcast:192.168.5.255  Mask:255.255.255.0
          inet6 addr: fe80::6ab5:99ff:fe78:ad6c/64 Scope:Link
          UP BROADCAST RUNNING MULTICAST  MTU:1500  Metric:1
          RX packets:4237972 errors:0 dropped:0 overruns:0 frame:0
          TX packets:2663665 errors:0 dropped:0 overruns:0 carrier:0
          collisions:0 txqueuelen:0
          RX bytes:547319641 (521.9 MiB)  TX bytes:220520225 (210.3 MiB)
 
lo        Link encap:Local Loopback
          inet addr:127.0.0.1  Mask:255.0.0.0
          inet6 addr: ::1/128 Scope:Host
          UP LOOPBACK RUNNING  MTU:16436  Metric:1
          RX packets:1090692 errors:0 dropped:0 overruns:0 frame:0
          TX packets:1090692 errors:0 dropped:0 overruns:0 carrier:0
          collisions:0 txqueuelen:0
          RX bytes:89713135 (85.5 MiB)  TX bytes:89713135 (85.5 MiB)
 
peth0     Link encap:Ethernet  HWaddr FE:FF:FF:FF:FF:FF
          inet6 addr: fe80::fcff:ffff:feff:ffff/64 Scope:Link
          UP BROADCAST RUNNING NOARP  MTU:1500  Metric:1
          RX packets:14557550 errors:102 dropped:0 overruns:0 frame:0
          TX packets:18532553 errors:0 dropped:0 overruns:0 carrier:0
          collisions:0 txqueuelen:1000
          RX bytes:2011028398 (1.8 GiB)  TX bytes:20743371955 (19.3 GiB)
 
tap0      Link encap:Ethernet  HWaddr FE:DC:03:FA:0A:FD
          inet6 addr: fe80::fcdc:3ff:fefa:afd/64 Scope:Link
          UP BROADCAST RUNNING MULTICAST  MTU:1500  Metric:1
          RX packets:0 errors:0 dropped:0 overruns:0 frame:0
          TX packets:340892 errors:0 dropped:0 overruns:0 carrier:0
          collisions:0 txqueuelen:500
          RX bytes:0 (0.0 b)  TX bytes:32530658 (31.0 MiB)
 
tap1      Link encap:Ethernet  HWaddr FE:93:1B:49:51:2B
          inet6 addr: fe80::fc93:1bff:fe49:512b/64 Scope:Link
          UP BROADCAST RUNNING MULTICAST  MTU:1500  Metric:1
          RX packets:0 errors:0 dropped:0 overruns:0 frame:0
          TX packets:340872 errors:0 dropped:0 overruns:0 carrier:0
          collisions:0 txqueuelen:500
          RX bytes:0 (0.0 b)  TX bytes:32529356 (31.0 MiB)
 
tap2      Link encap:Ethernet  HWaddr FE:E1:0F:73:C5:69
          inet6 addr: fe80::fce1:fff:fe73:c569/64 Scope:Link
          UP BROADCAST RUNNING MULTICAST  MTU:1500  Metric:1
          RX packets:0 errors:0 dropped:0 overruns:0 frame:0
          TX packets:340842 errors:0 dropped:0 overruns:0 carrier:0
          collisions:0 txqueuelen:500
          RX bytes:0 (0.0 b)  TX bytes:32527197 (31.0 MiB)
 
tap3      Link encap:Ethernet  HWaddr FE:CF:0C:41:CF:29
          inet6 addr: fe80::fccf:cff:fe41:cf29/64 Scope:Link
          UP BROADCAST RUNNING MULTICAST  MTU:1500  Metric:1
          RX packets:0 errors:0 dropped:0 overruns:0 frame:0
          TX packets:340839 errors:0 dropped:0 overruns:0 carrier:0
          collisions:0 txqueuelen:500
          RX bytes:0 (0.0 b)  TX bytes:32526987 (31.0 MiB)
 
vif0.0    Link encap:Ethernet  HWaddr FE:FF:FF:FF:FF:FF
          inet6 addr: fe80::fcff:ffff:feff:ffff/64 Scope:Link
          UP BROADCAST RUNNING NOARP  MTU:1500  Metric:1
          RX packets:2663690 errors:0 dropped:0 overruns:0 frame:0
          TX packets:4237985 errors:0 dropped:0 overruns:0 carrier:0
          collisions:0 txqueuelen:0
          RX bytes:220525759 (210.3 MiB)  TX bytes:547320421 (521.9 MiB)
 
vif1.0    Link encap:Ethernet  HWaddr FE:FF:FF:FF:FF:FF
          inet6 addr: fe80::fcff:ffff:feff:ffff/64 Scope:Link
          UP BROADCAST RUNNING NOARP  MTU:1500  Metric:1
          RX packets:402315 errors:0 dropped:0 overruns:0 frame:0
          TX packets:902391 errors:0 dropped:28279 overruns:0 carrier:0
          collisions:0 txqueuelen:500
          RX bytes:889703904 (848.4 MiB)  TX bytes:114468432 (109.1 MiB)
 
vif2.0    Link encap:Ethernet  HWaddr FE:FF:FF:FF:FF:FF
          inet6 addr: fe80::fcff:ffff:feff:ffff/64 Scope:Link
          UP BROADCAST RUNNING NOARP  MTU:1500  Metric:1
          RX packets:75264 errors:0 dropped:0 overruns:0 frame:0
          TX packets:397841 errors:0 dropped:28153 overruns:0 carrier:0
          collisions:0 txqueuelen:500
          RX bytes:196606268 (187.4 MiB)  TX bytes:76569944 (73.0 MiB)
 
vif3.0    Link encap:Ethernet  HWaddr FE:FF:FF:FF:FF:FF
          inet6 addr: fe80::fcff:ffff:feff:ffff/64 Scope:Link
          UP BROADCAST RUNNING NOARP  MTU:1500  Metric:1
          RX packets:6774649 errors:0 dropped:0 overruns:0 frame:0
          TX packets:9914065 errors:0 dropped:24570 overruns:0 carrier:0
          collisions:0 txqueuelen:500
          RX bytes:19334745479 (18.0 GiB)  TX bytes:1359548138 (1.2 GiB)
 
vif4.0    Link encap:Ethernet  HWaddr FE:FF:FF:FF:FF:FF
          inet6 addr: fe80::fcff:ffff:feff:ffff/64 Scope:Link
          UP BROADCAST RUNNING NOARP  MTU:1500  Metric:1
          RX packets:2416 errors:0 dropped:0 overruns:0 frame:0
          TX packets:341880 errors:0 dropped:0 overruns:0 carrier:0
          collisions:0 txqueuelen:500
          RX bytes:225475 (220.1 KiB)  TX bytes:32702082 (31.1 MiB)
 
virbr0    Link encap:Ethernet  HWaddr 00:00:00:00:00:00
          inet addr:192.168.122.1  Bcast:192.168.122.255  Mask:255.255.255.0
          inet6 addr: fe80::200:ff:fe00:0/64 Scope:Link
          UP BROADCAST RUNNING MULTICAST  MTU:1500  Metric:1
          RX packets:0 errors:0 dropped:0 overruns:0 frame:0
          TX packets:27 errors:0 dropped:0 overruns:0 carrier:0
          collisions:0 txqueuelen:0
          RX bytes:0 (0.0 b)  TX bytes:6748 (6.5 KiB)
 
xenbr0    Link encap:Ethernet  HWaddr FE:93:1B:49:51:2B
          UP BROADCAST RUNNING NOARP  MTU:1500  Metric:1
          RX packets:324811 errors:0 dropped:0 overruns:0 frame:0
          TX packets:0 errors:0 dropped:0 overruns:0 carrier:0
          collisions:0 txqueuelen:0
          RX bytes:26883663 (25.6 MiB)  TX bytes:0 (0.0 b)
 
[root@virtual2 pvasquez]#
 
Julio
 
 
Date: Sun, 29 Jul 2012 12:01:58 +0200
From: Mark van Dijk <lists+xen@xxxxxxxxxxxxxx>
To: xen-users@xxxxxxxxxxxxx
Subject: [Xen-users] Xen networking experiment (with custom scripts
        and     OpenVSwitch)
Message-ID: <20120729120158.73ab4f91@xxxxxxxxxxxxxx>
Content-Type: text/plain; charset=US-ASCII
 
Hello everyone,
 
Recently I have been testing my customized Xen 4.2 networking setup. It
works pretty good and I would like to share it with anyone who is
interested.
 
The relevant files can be found here:
https://github.com/slacks42/xenscripts
 
Benefit: configure the Xen related networking devices with one
understandable bash script.
 
Please note that this is all still work in progress. For example, some
logging entries should be deleted or modified, and some routines could
be cleaned up. Still, I think it's nice enough to share.
 
Description of the files:
 
xen.conf is basically a copy of /etc/xen/hotplugpath.sh and should be
put in /etc/xen/.
 
udev/xen-backend.rules is a modified version of the file supplied by
Xen. Line 6-9 take care of creation and deletion of
networking interfaces. As you can see, upon creation/deletion of a
network interface, the file /etc/xen/scripts/xennet is called. This
file can be found on github in the scripts directory.
 
Xennet is a replacement of vif-bridge. In the old scenario, udev
calls vif-setup, vif-setup calls vif-bridge, vif-bridge calls
vif-common and numerous other scripts. With xennet, I wanted to have one
script to take care of the networking. So xennet requires no other files
from /etc/xen/scripts.
 
As you can see xennet uses bash. It is not POSIX compliant but works
fine with bash. This takes us to the second millennium ;-) 
 
Here is a rundown of xennet:
 
Line 4 takes care of all error output from the script. This is a handy
way to debug, especially if you alter line 1 to '#!/bin/bash -x' so
that you can see exactly what the script does.
 
Then on line 8, the $unique variable is set to a random 6 character
wide string, a 'cookie' of sorts, this is used as a log prefix so that
you can see which particular instance of the script does what.
 
Then a couple of functions are initialized. The checklog() and
logmsg() functions take care of logging at the requested loglevel (set
in xen.conf). This is different from line 4; logmsg is a function that
logs to the console, syslog, or a file (xen.conf).
 
The sigerr(), fatal(), success(), xenstore_read_default(),
findCommand() and evalVariables() functions are modified versions of the
same functions in Xen's xen-script-common.sh file.
 
I added line 191-198 for debugging, rather than calling the actual
commands they fake their execution and only add a log entry. That's why
they are commented out.
 
The actual routine starts at line 200. Line 201 sets $command to
online/offline/add/remove depending on how the script was called from
xen-backend.rules. evalVariables (line 202) searches for arguments with
an '=', like 'foo=bar', and sets those variables accordingly (like
$foo == 'bar'). This is a nice trick I found in xen-script-common.sh.
 
On to line 220-292. This searches for the vifname and bridge name. If
$command is 'offline' or 'remove' then I found that it does not know
the vifname so it needs a way to find that. In all cases $vifname is
set to the requested vifname. Openvswitch does not require a bridge
name if you remove a device. So $bridge is not required with 'offline'
or 'remove'.
 
Line 297-337 adds or removes the vif from the switch. With openvswitch
this can be a "fake bridge", i.e. a VLAN tagged bridge, or an unmanaged
switch. One could easily replace the ovs-vswitch commands with brctl
commands if desired, I *think*.
 
Xen 4.2, when used with xl, does not setup or change your networking
(as we saw with older Xen and network-bridge). So you need to do that
yourself which is a much better idea imho anyway. In my case, my init
scripts start up openvswitch when the system boots and my custom
networking script creates the relevant switches and interfaces and
configures those.
 
Finally -- openvswitch can have a lot of messy output that can fill up
your syslog files. syslog-ng.conf is something I use to limit
openvswitch's output to /var/log/openvswitch.log.
 
Note that I seem to use the word 'switch' and 'bridge' while I am
talking about the same thing. Don't let this confuse you.
 
Comments are appreciated!
 
Mark
 
 
 
------------------------------
 
_______________________________________________
Xen-users mailing list
Xen-users@xxxxxxxxxxxxx
http://lists.xen.org/xen-users
 
 
End of Xen-users Digest, Vol 89, Issue 45
*****************************************
 
 
_______________________________________________
Xen-users mailing list
Xen-users@xxxxxxxxxxxxx
http://lists.xen.org/xen-users
_______________________________________________
Xen-users mailing list
Xen-users@xxxxxxxxxxxxx
http://lists.xen.org/xen-users

 


Rackspace

Lists.xenproject.org is hosted with RackSpace, monitoring our
servers 24x7x365 and backed by RackSpace's Fanatical Support®.