I am migrating a xen guest. It is running a CentOS 4. I am using the same steps that have worked in the past for our RHEL5 and RHEL6 based systems but for some reason, I’m getting errors.

Here are the steps in brief:

1. On the guest, install a non-xen kernel and the kernel’s devel package.
2. Turn off database and other services that may cause data inconsistencies
3. Using VMWare vCenter Converter, migrate data to VMWare
4. Boot into rescue mode using CentOS 4.8 CD disc 1
5. Change root to the OS’ root mount that rescue mode creates: chroot /mnt/sysimage
6. Modify /etc/grub/grub.conf so that the non-xen stanza is default
7. Remove non-xen modules from config: sed -i.pre-vmware '/xen/d' /etc/modprobe.conf
8. Check /etc/inittab and remove co runlevel and make sure we have lines for other runlevels
9. Double check /etc/inittab to make sure the Converter configured it correctly
10. Re-install grub: grub-install --recheck /dev/sda
11. Finally, recreate init ram disk: mkinitrd -v -f /boot/initr<kernel version>.img <kernel version>

This is what I get when I reboot now:

Can anyone help? Or provide some some troubleshooting tips? TIA.

There is a Red Hat KBase article discussing this issue and a solution. Essentially it looks like you do not have the right storage drivers for your disks. Be sure to include the scsi drivers when you make your initrd image:

# mkinitrd -f -v --with=scsi_mod --with=sd_mod --with=mptbase --with=mptscsi --with=mptspi --with=mptsas --with=mptscsih /boot/initrd-<version>.ELsmp.img <version>.ELsmp

Check more discussion of this question.

I try to boot a XEN VM and it looks like the hard drive naming changed from /dev/sda1 to /dev/hda1. So the boot process drops me to a busy box in initramfs. How can I specify the probably right boot device on the fly and continue the booting?

If you’re asking how to boot this one time, usually you just mount the filesystems to their correct locations, then exit and the boot process will continue.

If you want to fix this permanently you can change your fstab file to use drive labels.

Check more discussion of this question.

I’m following this construction to install (version 4.1.2) on a CentOS 6.2 (x86_64) system.
When I use this command

yum install http://au1.mirror.crc.id.au/repo/kernel-xen-release-6-3.noarch.rpm

I face this error:

Loaded plugins: fastestmirror, security
Loading mirror speeds from cached hostfile
 * base: ftp.sh.cvut.cz
 * extras: ftp.sh.cvut.cz
 * updates: ftp.sh.cvut.cz
Setting up Install Process
Cannot open: http://au1.mirror.crc.id.au/repo/kernel-xen-release-6-3.noarch.rpm. Skipping.
Error: Nothing to do

there is a third-party repository with some mirror list, but I don’t know if I could use them? and how would that possible?
I would appreciate if any one could lead me to a available link to download Xen Repository.

You received the error because the file does not exist.

The reason the file does not exist is that the file in the repo is named kernel-xen-release-6-4.noarch.rpm, while your howto has a 3 instead of the 4.

Check more discussion of this question.

I am having a server running XEN,with ubuntu 12.04 as Domain 0.

Intel VT is turned on by default and i dont want to run any Windows OS as VM.
So

1. how can i makesure that all the created VM will be in
paravirtualized mode even with VT enabled?

I have heard that in Full Virtualization hypervisor needs to work hard for emulating all the underlying hardware,which may reduce performance for too much i/o specific applications when compared to para mode.

2.Since i am running only Linux VMs do i really need to enable VT
support?3.Can i be able to run Linux VM in para mode with VT enabled?4.Will simply turning on VT support changes virtualization from para to full virtualization??5.Will there be any performance improvement for VM running Linux OS if VT is enabled?

1. how can i makesure that all the created VM will be inparavirtualized mode even with VT enabled?

when creating Vm specify whether its in para or full virtmode.
For vm to run in para mode it needs special kernel which need to be downloaded from online.So installing from CD may not work.

2.Since i am running only Linux VMs do i really need to enable VT support?

if running vm in para mode there is no need of VT.

3.Can i be able to run Linux VM in para mode with VT enabled?

Yes,Even with VT enabled Linux Vms can be run in para mode.

 4.Will simply turning on VT support changes virtualization from para to full virtualization??

No,it means we can run both para and full vms.We can choose whether to use vm in full mode or in para mode.

5.Will there be any performance improvement for VM running Linux OS if VT is enabled?

If VM is in para mode,enabling VT has no impact on VM.but HVM can be installed only with VT support.

Check more discussion of this question.

We’ve got a CentOS server running a cluster of virtuals. Occasionally the cluster’s internal network drops out for a minute or so … and then comes back. The problem is somehow related to the actual network traffic, but it is not a simple load issue. (The system is generally lightly loaded, and the problem occurs irrespective of actual load.)

The setup:

• CentOS 5.6 on Dom0, various CentOS on the DomU’s
• Hardware – a Dell R710 with a BroadCom NextXpress 2 NIC (sigh)
  • using the latest drivers for the NIC from BroadCom
• Xen configured to use network-bridge and vif-bridge
• Some iptable tweaks to route an unrelated port to one of the virtuals.

The system has one externally visible IP address, and Dom0 runs an Apache httpd configured with a number of virtual hosts each of which reverse proxies to web servers running on the virtuals. (The virtuals have to be NAT’ed, primarily because we don’t have enough allocated public IP addresses.)

The symptoms:

• Works fine most of the time.
• When someone tries to UPLOAD a large file to one virtuals, the internal network drops out … for all virtuals:
  • The Dom0 httpd sees a network timeout talking to the backend server on the virtual and reports a 502.
  • A previously established ssh connection from Dom0 to any of the DomU’s freezes.
  • Our monitoring shows ping failures for traffic between virtuals.
  • The Xen consoles to the DomU’s do not freeze.
  • No log messages in any log files that I can see, on either Dom0 or the DomU’s … apart from the Dom0 httpd logs.
  • After a minute or so, the problem clears by itself.

This is 100% reproducible.

What we’ve tried:

• Downloading, building and installing the latest BNX2 driver on Dom0
• Turning off MSI on the NIC – adding “options bnx2 disable_msi=1″ to /etc/modprobe.conf
• Turning off tcp segmentation offload – “ethtool -K eth0 tso off”.
• Sacrificing a black rooster at midnight.

I’ve exhausted all my options apart from switching to KVM … or slaughtering more roosters.

Any suggestions?

We eventually did find the problem. It turned out to be a caused by a problem in our virtual network configuration. For some reason that I can no longer remember exactly, network traffic for that particular download was taking an extra loop through the virtual networks. When a user tried to upload a large file, the download was tying down all available kernel network buffers. That was causing the entire network to freeze … until something timed out and it all unjammed.

I’m sorry that this is all a bit vague, but there may offer some hints for people who run into a similar problem.

Check more discussion of this question.

Can xen server be configured to address more than 4Gb RAM.
Since its based on a 32 bit kernel,is it possible to address more than 4GB ram.I have heard about PAE can it solve this?

Will changing the default 750M to more than 4Gb results in any performance degradation?

The only reason i want to add more ram to host is to load a 5 GB VM to completely run on RAM,for maximum performance benefits,for some testing purpose.

Installed RAM is more than 64 GB,and a lot of unused RAM.

With a 3-2bit System, every application can address up to 4GB of memory. PAE only lets you put more RAM in the Server to let more applications run with their 4GB.

If you really need to address more then 4GB you have to switch to a 64-bit kernel.

Edit

This document suggests that Xen Cloud Platform is based on a 64-bit system:

The Xen Cloud Platform host is a 64-bit x86 server-class machine
devoted to hosting mul-tiple VMs. This machine runs a stripped-down
Linux operating system with a Xen-enabled kernel which controls the
interaction between the virtualized devices seen by VMs and the
physical hardware.

Check more discussion of this question.

i want to create a tmpfs file system for one of my vm to run on,using XEN Server.

On the host machine i tried this but is not working,.

xe sr-create device-config:device=/dev/shm name-label=”RAM”
type=tmpfs

Is it possible to make a tmpfs file system using xe command and make the vm run on it rather on HDD.This VM should not be slowed down by waiting for disks(This is what required).Can this be achieved using XEN server.Is there a better way of doing it?
This host is having only a single HDD and can not afford to add extra storage.

I’m guessing here, but tmpfs generates a filesystem. SR’s are block devices. You could probably create an empty file on the tmpfs, run losetup on it, and use that as your block device on tmpfs.

Edit:
I just did a quick test:

dd if=/dev/zero out=/dev/shm/testing bs=4M count=100
losetup /dev/loop0 /dev/shm/testing
parted /dev/loop0 mklabel msdos
parted /dev/loop0 mkpart 1 primary ext2 0% 100%
pvcreate /dev/loop0

At this moment you can make it into a part of your LVM, so you should be able to put images on it.

Check more discussion of this question.

I have been using virt manager to manage XEN and it crashes often so i moved to Citrix Xencenter but i can not make it work.

Details:
using XEN 4.1 running on ubuntu 12.04

citrix XenCenter Installed on windows XP

No firewall or network issue exist as i can ping XEN running on ubuntu 12.04.

when i try to add server its showing

Unable to Connect to server ‘ip-address’ connection was refused.Check
that XENServer is configured correctly on ‘ip-address’ and try again

Should i need to enable anything in XEN Configuration file to allow accessing XEN from citrix XENCenter..?

Citrix Xen is a complete operating system on it’s own running from a modified CentOS machine. If you are not running Citrix Xen on your server, then you will not be able to connect to it with XenCenter.

You can not use XenCenter with a default Xen, your OS has to be XenServer.

You can get it over here: http://www.citrix.com/lang/English/lp/lp_2304908.asp

Check more discussion of this question.

Can we create and run a virtual machine that uses RAM only or minimized HDD usage.

That is same as the working concept as that of live CD.

I know It is possible,but still confused,whether it can be implemented.

If you have sufficient ram, create a large tmpfs ram drive and put your hard drive images there.

Just like a live cd things will be lost when the system crashes or reboots. You will probably need scripts that sync it to a real drive on a regular basis and before shutting down.

Why are you trying to minimize hard drive use? That might help figure out the best solution.

Check more discussion of this question.

I have a problem with the Windows Server 2008 guest (hvm). I can’t get a network interface running for him.

I also have a Debian guest and it’s working ok, but I can’t do it with the Win2k8 guest. When I started the VM, the machine freezes and I can’t connect by ssh to the host.

/etc/network/interfaces

# The loopback network interface
auto lo
iface lo inet loopbackauto eth0
iface eth0 inet static
        address 188.165.B.C
        netmask 255.255.255.0
        network 188.165.B.0
        broadcast 188.165.255.255
        gateway 188.165.B.254

brctl show

bridge name bridge id       STP enabled interfaces
eth0        8000.e840f20acc28   no      peth0

/etc/xen/xend-config.sxp

...
(vif-script vif-bridge)
(network-script 'network-bridge') 
...

/etc/xen/win2k8.cfg

#  Networking
#
vif         = [ 'ip=5.39.F.G,mac=yy:yy:yy:yy:yy:yy,type=ioemu,bridge=eth0' ]

/etc/xen/debian.cfg

#  Networking
#
vif         = [ 'ip=178.33.D.E,mac=xx:xx:xx:xx:xx:xx' ]

As you can see, in the Debian guest I only have to specify an IP address and a MAC. But if I put that in the Win2k8 guest, the machine does not start.

I am using Xen 4.0

I am surprised to see “peth0″ in your btctl show output. Perhaps your issue is a simple typo? (but then no idea how your Debian VM should work)

And in my only Debian based system I have right now with a bridge (A proxmox 2.1 machine), I also have the bridge in /etc/network/interfaces:

# cat /etc/network/interfaces
[...]
iface eth0 inet manualiface eth1 inet manualiface eth2 inet manualiface eth3 inet manualauto vmbr0
iface vmbr0 inet static
    address  10.11.12.13
    netmask  255.255.0.0
    gateway  10.11.0.1
    bridge_ports eth0 eth1 eth2 eth3
    bridge_stp off
    bridge_fd 0
[...]# brctl show
bridge name     bridge id               STP enabled     interfaces
vmbr0           8000.0025903b7d66       no              eth0
                                                        eth1
                                                        eth2
                                                        eth3
                                                        tap100i0

Check more discussion of this question.