Multiple virtual to physical nic setup docs
by Alex.Upton@instinet.com
All,
I'm looking for documentation that can tell me how I can do the following
with xen.
basically I have a dual homed system, and want all virtual eth0 nics to
utilize the real physical eth0 nic, and would like all virtual eth1 nics
to use the real physical eth1 nic.
as of now I have this:
in my /etc/xen/host1
vif = [ 'mac=0a:00:00:00:00:01',
'mac=0b:00:00:00:00:01']
in my /etc/xen/xend-config.sxp
(network-script 'network-bridge bridge=br0 netdev=eth0')
(network-script 'network-bridge bridge=br1 netdev=eth1')
on the virtual host i ssh in through the virtual eth0 interface
then when i issue a tcpdump on the virtual eth1 i see real traffic for the
real physical eth1, however when assigning an address on the virtual eth1,
i can't ping it from outside the virtual environment.
Any direction to specific setup docs would be greatly appreciated.
Thanks,
-Alex
*****************************************************************
<<<Disclaimer>>>
In compliance with applicable rules and regulations, Instinet
reviews and archives incoming and outgoing email communications,
copies of which may be produced at the request of regulators.
This message is intended only for the personal and confidential
use of the recipients named above. If the reader of this email
is not the intended recipient, you have received this email in
error and any review, dissemination, distribution or copying is
strictly prohibited. If you have received this email in error,
please notify the sender immediately by return email and
permanently delete the copy you received.
Instinet accepts no liability for any content contained in the
email, or any errors or omissions arising as a result of email
transmission. Any opinions contained in this email constitute
the sender's best judgment at this time and are subject to change
without notice. Instinet does not make recommendations of a
particular security and the information contained in this email
should not be considered as a recommendation, an offer or a
solicitation of an offer to buy and sell securities.
*****************************************************************
18 years, 1 month
RE: [Fedora-xen] two fully virtualized systems crashing
by Matthias Bayer
Hi Stephen,
thanks a lot for sharing this information - very helpful !
i will shortly give it a try.
thanks,
Matthias
> -----Ursprüngliche Nachricht-----
> Von: "Stephen C. Tweedie" <sct(a)redhat.com>
> Gesendet: 12.04.06 16:43:32
> An: Alex.Upton(a)instinet.com
> CC: fedora-xen(a)redhat.com
> Betreff: RE: [Fedora-xen] two fully virtualized systems crashing
> Hi,
>
> On Wed, 2006-04-12 at 08:18 -0400, Alex.Upton(a)instinet.com wrote:
> > Sorry, black screen meaning that all VGA output is disabled.
> >
> > And generally if there's a kernel panic or something core dumps there's a
> > log to trace the errors back. In this scenario there isn't.
>
> In that sort of situation, serial console is often the best way to try
> to capture information when using Xen. I've just added a section on it
> to the FC5 Xen wiki:
>
> http://fedoraproject.org/wiki/FedoraXenQuickstartFC5
>
> which should be helpful in setting this up.
>
> Thanks,
> Stephen
>
>
> --
> Fedora-xen mailing list
> Fedora-xen(a)redhat.com
> https://www.redhat.com/mailman/listinfo/fedora-xen
_______________________________________________________________
SMS schreiben mit WEB.DE FreeMail - einfach, schnell und
kostenguenstig. Jetzt gleich testen! http://f.web.de/?mc=021192
18 years, 1 month
RE: [Fedora-xen] two fully virtualized systems crashing
by Alex.Upton@instinet.com
Ahh very helpful!
Most appreciated, Thanks!
Don't suppose you have any docs or know where I can read up on multi-nic
setup? The documentaton on the official xen wiki leaves much to be
desired.
-Alex
-----Original Message-----
From: Stephen C. Tweedie [mailto:sct@redhat.com]
Sent: Wednesday, April 12, 2006 10:43 AM
To: Alex Upton
Cc: fedora-xen(a)redhat.com; sct(a)redhat.com
Subject: RE: [Fedora-xen] two fully virtualized systems crashing
Hi,
On Wed, 2006-04-12 at 08:18 -0400, Alex.Upton(a)instinet.com wrote:
> Sorry, black screen meaning that all VGA output is disabled.
>
> And generally if there's a kernel panic or something core dumps
> there's a log to trace the errors back. In this scenario there isn't.
In that sort of situation, serial console is often the best way to try to
capture information when using Xen. I've just added a section on it to
the FC5 Xen wiki:
http://fedoraproject.org/wiki/FedoraXenQuickstartFC5
which should be helpful in setting this up.
Thanks,
Stephen
*****************************************************************
<<<Disclaimer>>>
In compliance with applicable rules and regulations, Instinet
reviews and archives incoming and outgoing email communications,
copies of which may be produced at the request of regulators.
This message is intended only for the personal and confidential
use of the recipients named above. If the reader of this email
is not the intended recipient, you have received this email in
error and any review, dissemination, distribution or copying is
strictly prohibited. If you have received this email in error,
please notify the sender immediately by return email and
permanently delete the copy you received.
Instinet accepts no liability for any content contained in the
email, or any errors or omissions arising as a result of email
transmission. Any opinions contained in this email constitute
the sender's best judgment at this time and are subject to change
without notice. Instinet does not make recommendations of a
particular security and the information contained in this email
should not be considered as a recommendation, an offer or a
solicitation of an offer to buy and sell securities.
*****************************************************************
18 years, 1 month
RE: [Fedora-xen] two fully virtualized systems crashing
by Alex.Upton@instinet.com
Sorry, black screen meaning that all VGA output is disabled.
And generally if there's a kernel panic or something core dumps there's a
log to trace the errors back. In this scenario there isn't.
And I should clarify the reboot as well, its not a software initiated
reboot, it's a hard reboot.
Admittedly very odd!
Thanks,
-Alex
-----Original Message-----
From: fedora-xen-bounces(a)redhat.com [mailto:fedora-xen-bounces@redhat.com]
On Behalf Of Stephen C. Tweedie
Sent: Tuesday, April 11, 2006 11:17 PM
To: Alex Upton
Cc: fedora-xen(a)redhat.com
Subject: RE: [Fedora-xen] two fully virtualized systems crashing
Hi,
On Tue, 2006-04-11 at 11:24 -0400, Alex.Upton(a)instinet.com wrote:
> If I'm copying data to or from a virtualized host the parent system
> will black screen without any kernel dump info, thus all i/o is halted
> and the only solution is a reboot.
>
> Any thoughts?
The original problem described a spontaneous reboot, not a "black screen"
(what exactly do you mean by that, btw? The screen saver kicks
in?) And what do you mean by kernel dump info --- HV VGA output, dmesg
logs, or what?
Thanks,
Stephen
--
Fedora-xen mailing list
Fedora-xen(a)redhat.com
https://www.redhat.com/mailman/listinfo/fedora-xen
*****************************************************************
<<<Disclaimer>>>
In compliance with applicable rules and regulations, Instinet
reviews and archives incoming and outgoing email communications,
copies of which may be produced at the request of regulators.
This message is intended only for the personal and confidential
use of the recipients named above. If the reader of this email
is not the intended recipient, you have received this email in
error and any review, dissemination, distribution or copying is
strictly prohibited. If you have received this email in error,
please notify the sender immediately by return email and
permanently delete the copy you received.
Instinet accepts no liability for any content contained in the
email, or any errors or omissions arising as a result of email
transmission. Any opinions contained in this email constitute
the sender's best judgment at this time and are subject to change
without notice. Instinet does not make recommendations of a
particular security and the information contained in this email
should not be considered as a recommendation, an offer or a
solicitation of an offer to buy and sell securities.
*****************************************************************
18 years, 1 month
RE: [Fedora-xen] two fully virtualized systems crashing
by Alex.Upton@instinet.com
I have a similar issue on a Dell Power Edge 1850 running:
Fedora Core 5
2.6.16-1.2080_FC5xen0
xen-3.0.1-4
Except my issue appears to affect the parent host as well. It seems to be
relevant during any increased NIC utilization.
If I'm copying data to or from a virtualized host the parent system will
black screen without any kernel dump info, thus all i/o is halted and the
only solution is a reboot.
Any thoughts?
-----Original Message-----
From: fedora-xen-bounces(a)redhat.com [mailto:fedora-xen-bounces@redhat.com]
On Behalf Of Stephen C. Tweedie
Sent: Tuesday, April 11, 2006 11:18 AM
To: matbayer.web(a)web.de
Cc: fedora-xen(a)redhat.com
Subject: Re: [Fedora-xen] two fully virtualized systems crashing
Hi,
On Tue, 2006-04-11 at 16:05 +0200, Matthias Bayer wrote:
> i have serveral DomU's running under Xen xen-3.0.1-4
kernel-xen0-2.6.16-1.2080_FC5.
> i have paravirtualized and fully virtualized systems (windows and
linux).
> now the problem is whenever i start 2 fully virtualized systems in
parallel (first start one then if this runs try to start the other one)
the computer completely crashes and reboots, no more log entries...
nothing.
>
> does anybody now how to debug such a problem ?
With serial console, usually. There are also a number of upstream Xen
updates for VMX, and we're putting together an update which may help
there.
Cheers,
Stephen
--
Fedora-xen mailing list
Fedora-xen(a)redhat.com
https://www.redhat.com/mailman/listinfo/fedora-xen
*****************************************************************
<<<Disclaimer>>>
In compliance with applicable rules and regulations, Instinet
reviews and archives incoming and outgoing email communications,
copies of which may be produced at the request of regulators.
This message is intended only for the personal and confidential
use of the recipients named above. If the reader of this email
is not the intended recipient, you have received this email in
error and any review, dissemination, distribution or copying is
strictly prohibited. If you have received this email in error,
please notify the sender immediately by return email and
permanently delete the copy you received.
Instinet accepts no liability for any content contained in the
email, or any errors or omissions arising as a result of email
transmission. Any opinions contained in this email constitute
the sender's best judgment at this time and are subject to change
without notice. Instinet does not make recommendations of a
particular security and the information contained in this email
should not be considered as a recommendation, an offer or a
solicitation of an offer to buy and sell securities.
*****************************************************************
18 years, 1 month
eth0 not coming up in domU
by Duncan Morgan
Hello,
I have a FC5 dom0 and have set up a White Box Linux 3.2 domU and the eth0
will not come up in this domain. I see the following error message
ifup eth0
modprobe: QM_MODULES: Function not implemented
modprobe: QM_MODULES: Function not implemented
e100 device eth0 does not seem to be present, delaying initialization.
The peth0 and xenbr0 are running in dom0 and appear to be configured
correctly.
I understand that the error message is about modules.conf versus
modprobe.com but I'm told that this error message is irrelevant since the
unprivileged kernel does not load modules.
Thanks in advance for any help with this?
Duncan
18 years, 1 month
Tracebacks
by Ted Kaczmarek
When try to use either sda1/sda2 sdb1/sdb2 root/swap combo I get
tracebacks when launching domu.
netfront: Initialising virtual ethernet driver.
Registering block device major 8
register_blkdev: cannot get major 8 for sd
xen_blk: can't get major 8 with name sd
NET: Registered protocol family 2
Unable to handle kernel NULL pointer dereference at virtual address
00000104
printing eip:
c0222268
*pde = ma 00000000 pa 55555000
Oops: 0002 [#1]
SMP
Modules linked in:
CPU: 0
EIP: 0061:[<c0222268>] Not tainted VLI
EFLAGS: 00010086 (2.6.12.6-xenU)
EIP is at blk_start_queue+0x8/0x60
eax: 00000000 ebx: 00000000 ecx: fbf8c000 edx: 00000000
esi: cbc24aa0 edi: c1278000 ebp: c1279f9c esp: c1279f30
ds: 007b es: 007b ss: 0069
Process xenwatch (pid: 7, threadinfo=c1278000 task=c1226560)
Stack: 00000004 cbc08000 c023507e 00000000 cbc08000 c0234f98 cbc08000
00000000
00000004 00000000 00000200 cbc08000 c02c50a0 c1279f78
c02ce518 c02c3a9f
c1279f7c 00000000 00000000 00000200 00b00000 c02333e0
c02327a7 cbc08000
Call Trace:
[<c023507e>] kick_pending_request_queues+0x1e/0x40
[<c0234f98>] connect+0xe8/0x130
[<c02333e0>] otherend_changed+0x0/0xa0
[<c02327a7>] xenwatch_thread+0x87/0x140
[<c01353d0>] autoremove_wake_function+0x0/0x60
[<c0232720>] xenwatch_thread+0x0/0x140
[<c0134ee7>] kthread+0xc7/0xd0
[<c0134e20>] kthread+0x0/0xd0
[<c01077fd>] kernel_thread_helper+0x5/0x18
Code: 00 00 00 8b 44 24 04 8b 48 60 ff e1 8d b4 26 00 00 00 00 81 44
24 04 a0 00 00 00 e9 d3 25 00 00 8d 76 00 53 83 ec 04 8b 5c 24 0c
<f0> 0f ba b3 04 01 00 00 02 f0 0f ba ab 04 01 00 00 06 19 c0 85
<6>IP: routing cache hash table of 2048 buckets, 16Kbytes
I switched to using hda1/hda2 combo and the domU came up. This may be
particular to my mobo setup, have seen this type of problem on and
off since testing the hg stuff before 3.0 was released.
disk = [ 'phy:VolGroup01/debian,hda1,w' , 'phy:VolGroup01/
debswap,hda2,w']
The part that really messes with me is the domU is a debian image
from jailtime.org and its fstab is
dev/sdd1 / ext3 errors=remount-ro 0 1
/dev/sdd2 none swap sw 0 0
proc /proc proc defaults 0 0
I am a tad out of the loop, but from what I recall in the past the
fstab had to match what was provided in the domU's config.
kernel-xenU-2.6.16-1.2080_FC5
kernel-xen0-2.6.16-1.2080_FC5
This also was a yum upgrade.
Should I bugzilla this or is the target still moving :-)
Regards,
Ted
18 years, 1 month
need advices on how to install domU with xenguest-install
by Denis Forveille
After reading instructions from from the "FedoraXenQuickstartFC5" doc
and the xen user manual, I'm using the xenguest-install script to
install my domUs. However, I have some questions on how you guys,
perform installations of your domU servers:
This is what I'm doing:
- on dom0, create one lvg per domU (/dev/vg/domU1,/dev/vg/domU2,etc..)
- format the lvg with "mkfs -t ext3 /dev/vg/domU1"
- run xenguest-install, answering "/dev/vg/domU1" to the "What would
you like to use as the disk (path)" question
The goals is to be able later to mount that lvg and copy files from/to
it (when the associated domU is down of course...) and later to clone it
to create other domUs from it..
Anaconda starts fine, but I always get a message saying that the file
system is not valid and it has to be reinitialised...
Q: Is that normal? "/dev/vg/domU1" IS correct and the fs IS correct
before running xenguest-install (checked by mount/copy/umount..)
Then with diskdruid, I create only one partition in xvda (mount="/",
fs=ext3).
Anaconda gives a warning saying that there is no swap partition defined...
Q: Is there a way to run xenguest-install and specify another lvg for
the swap partition?
Q: What guy do you do? do You define a swap partition? do you initially
not define a swap partition and then, when he domU is installed, modify
the domU configuration file to add another lvg for the swap, start the
domU and configure there the swap partition?
(It would be possible with disk druid to build another structure in the
lvg (with "/boot", "/" and swap partition...) but I'll not be able to
use the lvg from dom0 later, I guess. Remember, i want to be able to
mount the lvg from dom0 when the domU is down)
Then comes grub...
- When anaconda ask for it, how do you configure grub on domU. Do you:
1) install it on the MBR of first disk (MBR of the disk with is mapped
to the lvg?)
2) install it in first block of first partition (see remark above..)
3) don't install it, and after installation of domU. copy the kernel
in /boot of dom0, update the domU configuration file to set the kernel
and root parameter
It seems that option 3 is the only one if I want to be able to mount
the lvg from dom0 later...
Q: what guys do you recommend there?
I also tried to install FC5 by booting from the DVD on an lvg directly
(in "/dev/vg/domU1"), but was not able to start it as a domU...
Installation is OK, I used the existing "/boot" partition when asked for
it (the one used by dom0)and tried to create a domU. My domU started but
I had a lot of problems.. Most of them because I didn't manage to
install FC5 in the lvg with the FC5xenU kernel (I checked all the menus
where packages and selected and dint find any entry to choose the
kernel). The standard kerne lwas installed. I successfully boot the domU
with the xenU kernel (parameter kernel pointing to it in /boot in domu
config file), but in domU there is no corresponding /lib/module dir...
and because of that, thenetwork does not start ..(I stop here..)
Q: is there a way to install FC5 this way (boot from DVD) and force it
to install the FC5xenU kernel instead of the standard FC5 kernel?
MANY MANY thaks for sharing.
(Sorry for the loooonnng post..)
18 years, 1 month
two fully virtualized systems crashing
by Matthias Bayer
Hi,
i have serveral DomU's running under Xen xen-3.0.1-4 kernel-xen0-2.6.16-1.2080_FC5.
i have paravirtualized and fully virtualized systems (windows and linux).
now the problem is whenever i start 2 fully virtualized systems in parallel (first start one then if this runs try to start the other one) the computer completely crashes and reboots, no more log entries... nothing.
does anybody now how to debug such a problem ?
for couriosity this occurs only on one system, i have a second system with a very similar setup and here i can start 2 fully virtualized (windows) systems in parallel without problem.
any ideas ?
regards,
Matthias
_______________________________________________________________
SMS schreiben mit WEB.DE FreeMail - einfach, schnell und
kostenguenstig. Jetzt gleich testen! http://f.web.de/?mc=021192
18 years, 1 month
RE: [Fedora-xen] Error creating Domain
by You, Yongkang
What's your configuration for xenU?
Did you give a right disk option?
Best Regards,
Yongkang (Kangkang) 永康
>-----Original Message-----
>From: fedora-xen-bounces(a)redhat.com
>[mailto:fedora-xen-bounces@redhat.com] On Behalf Of shyamal thatte
>Sent: 2006年4月11日 15:51
>To: fedora-xen(a)redhat.com
>Subject: [Fedora-xen] Error creating Domain
>
>Hi,
>
>Need help please
>
>XEN 3.0.1 on FC4 is domain 0.
>
>Installed FC4 on /dev/hda8 separate installation.
>
>When trying to 'xm create ' fails with error 0
>There does not seem to be a Xen 3 User kernel. Can one create it ?
>
>Attached please find the xend logs and configuration file
>
>Thanks in advance for your help.
>
>Regards
>Shyamal
18 years, 1 month