virt January 2008

virt@lists.centos.org

19 participants
14 discussions

Kernel panic in DomU when using swiotlb=force option
by Michael Olson 29 Jan '08

29 Jan '08

I'm trying to get a PVR-150 PCI card imported into a Xen DomU, however the driver blows up the kernel when loaded. The work around for this mentioned in various places is to use swiotlb=force on the DomU kernel. That leads to a fast and furious kernel panic. I've tried this with all Dom0 & DomU running kernel-xen-2.6.18-53.el5 kernel-xen-2.6.18-53.1.6.el5, bot x86 and 86_64 versions, also tried iommu=soft and various sizing options with swiotbl. Any ideas on how to fix or work-around this? Thanks -- Michael Kernel splat: ------------------------------------------------------- [root@hydra mythic]# xm create mythic -c Using config file "./mythic". Started domain mythic Bootdata ok (command line is ro root=LABEL=/1 swiotlb=force) Linux version 2.6.18-53.1.6.el5xen (mockbuild(a)builder6.centos.org) (gcc version 4.1.2 20070626 (Red Hat 4.1.2-14)) #1 SMP Wed Jan 23 11:59:21 EST 2008 BIOS-provided physical RAM map: Xen: 0000000000000000 - 0000000030800000 (usable) No mptable found. Built 1 zonelists. Total pages: 198656 Kernel command line: ro root=LABEL=/1 swiotlb=force Initializing CPU#0 PID hash table entries: 4096 (order: 12, 32768 bytes) Xen reported: 2400.082 MHz processor. Console: colour dummy device 80x25 Dentry cache hash table entries: 131072 (order: 8, 1048576 bytes) Inode-cache hash table entries: 65536 (order: 7, 524288 bytes) ----------- [cut here ] --------- [please bite here ] --------- Kernel BUG at lib/../arch/i386/kernel/swiotlb.c:160 invalid opcode: 0000 [1] SMP last sysfs file: CPU 0 Modules linked in: Pid: 0, comm: swapper Not tainted 2.6.18-53.1.6.el5xen #1 RIP: e030:[<ffffffff80334da7>] [<ffffffff80334da7>] swiotlb_init_with_default_size+0xa0/0x19e RSP: e02b:ffffffff805d9f40 EFLAGS: 00010282 RAX: 00000000fffffff4 RBX: 0000000000000000 RCX: 0000000000063906 RDX: ffffffffff578000 RSI: 0000000000000000 RDI: 0000000000000000 RBP: 0000000004000000 R08: 0000000000001000 R09: ffffffff806399e0 R10: 0000000000000000 R11: 0000000000000000 R12: 0000000000000000 R13: 0000000000000000 R14: 0000000000000000 R15: 0000000000000000 FS: 0000000000000000(0000) GS:ffffffff80599000(0000) knlGS:0000000000000000 CS: e033 DS: 0000 ES: 0000 Process swapper (pid: 0, threadinfo ffffffff805d8000, task ffffffff804c4a00) Stack: 0000000000000295 0000000000020800 0000000000000000 ffffffff80273167 0000000000000000 ffffffff805ec50a 0000000000000000 0000000000020800 0000000000000000 0000000000000000 Call Trace: [<ffffffff80273167>] pci_swiotlb_init+0x9/0x2d [<ffffffff805ec50a>] mem_init+0x5e/0x233 [<ffffffff805e2a7a>] start_kernel+0x189/0x224 [<ffffffff805e21ed>] _sinittext+0x1ed/0x1f3 Code: 0f 0b 68 1b e1 47 80 c2 a0 00 48 83 eb 80 48 8b 05 4c ca 33 RIP [<ffffffff80334da7>] swiotlb_init_with_default_size+0xa0/0x19e RSP <ffffffff805d9f40> <0>Kernel panic - not syncing: Fatal exception -------------------------------------------------------

2 1

Network problems with Centos 5 xen kernel
by Richard Chapman 28 Jan '08

28 Jan '08

Hi - I have a machine which was originally built with Centos 5.0 including xen. It has all the current updates and here are some questions: The current problem is: 1) If I boot the non-xen kernel - it runs fine. If I boot the xen kernel - it boots - but has no network connectivity - and I get the following errors continuously in the messages log: Time/date: C5 Kernel: vif0.0: received packet with own address as source address. And another question: 2) The Centos 5.1 release notes say that the 5.1 xen release includes dnsmasq, and that this may conflict with dhcpd and named if you are running these. I am running both - and i thought dnsmasq was not running because it is set to "not start at boot" - but I now notice that it is "running" even though it is set to "not start on boot". This is true whether I boot the xen or non-xen kernel. Is this a problem - and if so - how do I stop it from running? How do dnsmasq and dhcpd both function without breaking each other? Now some detailed background: The machine uses a Realtek 8111B network adaptor - and this has been a problem all along. I used to manually reinstall the r1000 driver (apparently the correct driver for the adaptor) every time the kernel was updated. Tedious - but it worked. My xen kernel worked fine. Some months back - a bunch of updates came through corresponding to the release of Centos 5.1 - and my normal driver re-install didn't seem to get the Xen kernel running. On advice from the Centos forums - I installed the dkms-r1000 driver - and eventually got the non-xen kernel running fine - but couldn't get the xen kernel running. Every boot - the message "starting eth0 - Failed" would come up. I have been using the non-xen kernel ever since. My system has both kernel-devel and kernel-xen- devel installed. Just yesterday - I installed the latest Kernels ( kernel 2.6.18-53.1.6.el5 <https://192.168.0.201:10000/software/edit_pack.cgi?search=kernel&package=ke…>) - using gnome software updater which installs both xen and non-xen kernels. The non-xen kernel worked fine without any manual driver install. Presumably the dkms install worked fine. When I tried to boot the xen kernel - there were lots of problems - partly because I recently got the shcpd going on the server - and it now needs static ips. The xen kernel still seemed to be set to dynamic ips - but that was easily fixed. Now the message "starting etho - OK" is displayed - and nearly everything looks OK in the startup messages - but several things are slow to start. Once it boots - there is no network connectivity - and the above "vif0.0" errors appear in the messages log. Can anyone shed any light on these questions? Thanks Richard.

1 0

BUG: soft lockup detected on CPU#?
by Brett Worth 22 Jan '08

22 Jan '08

Hello All. I've just started looking into Xen and have a test environment in place. I'm seeing an annoying problem that I thought worthy of a post. Config: I have 2 x HP DL585 servers each with 4 Dual core Opterons (non-vmx) and 16GB RAM configured as Xen servers. These run CentOS 5.1 with the latest updates applied. These system both attach to an iSCSI target which is an HP DL385 running ietd and serving SAN based storage. I have a test VM running CentOS 5.1 also updated. Problem: If I run the VM on a single server everything is OK. If I do a migrate of the VM to the other server I start getting random "BUG: soft lockup detected on CPU#?" messages on the VM console. The messages seem to happen with IO but not every time. A reboot of the VM on the new server will stop these messages. I've also left the VM running overnight a couple of times and when I do I find that any external sessions (ssh) are hung in the morning but the console session is not. New ssh sessions can be started and seem to work. After much googling it looks like the kernel messages can occur if dom0 is very busy but mine is not. Any suggestions? Regards Brett Worth

3 4

Problem with CentOS 5.1 / Xen / Migration : domU time
by Frederic SOULIER 21 Jan '08

21 Jan '08

Hi all, I'm having issue with my Centos 5.1 /Xen installation. I'm having 2 dom0 running 2.6.18-53.1.4.el5xen (x86_64). These two dom0 are connected to a SAN. These two dom0 hosts multiple domU like Centos 4.5 (2.6.9-55.0.9.ELxenU) and Centos 5.1 (2.6.18-53.1.4.el5xen). These two dom0 run ntpd to synchronize dom0 time. When i start a virtual machine on the first dom0 all seems good, the domU time is like the dom0 time. The problem appears when i migrate the domU to another dom0. After the migration, the time in domU on the new dom0 is totally false (we can have more than one day of desynchronisation) and the time on the domU seems to be blocked (the clock indicates always the same time ). In the xend Logs we can't see any error message with the debug mode. The solution for this moment is to restart the domU in the new dom0 to be synchronized again. We manage a lot of authentications services like CAS or Shibboleth which uses timestamp for security validation and this problem is very annoying for us. Anyone has encountered this problem ? Any idea how to solve it ? Thanks for your help. Regards Fred -- Frederic Soulier Administrateur Reseau Centre de Ressources Informatiques Université Toulouse 1 Sciences Sociales 2 RUE DU DOYEN GABRIEL MARTY 31042 TOULOUSE CEDEX 9 Tel : +33 5 61 63 39 98 / +33 5 61 63 39 98 Fax : +33 5 61 63 37 98 Bureau : AR38 bis http://cri.univ-tlse1.fr/

2 2

Centos 4 physical to virtual VMWare
by Ted Miller 14 Jan '08

14 Jan '08

Any tips on taking a Centos 4 desktop from physical hardware to virtual environment? Ted Miller Indiana, USA

2 1

Firewalling domU
by Mindaugas 11 Jan '08

11 Jan '08

Hello, I think I saw this somewhere but cannot find it with google now. How to setup anti-spoofing firewall rules in dom0 so that domU could not change IP and try to get spoof other traffic? I could use physdev module of iptables but how to determine which vif belongs to which domU without too much scripting? Or parsing "xm list" and matching with some configuration file (like: domain1:192.168.1.10 domain2:192.168.1.22) is the way? Then how to automatically run the script after "xm create"? Thanks for the ideas. Question should be simple but I'm stuck here now. :) Mindaugas

1 0

[Fwd: [CentOS-devel] Preview Xen 3.2 rc* packages]
by Karanbir Singh 10 Jan '08

10 Jan '08

FYI -- Karanbir Singh : http://www.karan.org/ : 2522219@icq

1 0

Rebooting xen after CentOS update
by Kai Schaetzl 10 Jan '08

10 Jan '08

I updated a CentOS 5.0 Xen VM to 5.1 (host had been updated already last year) and then rebooted from within. This hung the machine on shutdown with nearly 100% CPU usage. I couldn't shut it down any other way than destroying it. After starting up it would seem to start up, but wasn't available via network or console. Shutting it down would display the same symptoms as before. Finally, I rebooted the host machine and then was able to start the updated guest as well. This isn't normal, is it? (This was my first number update within a Xen VM.) Kai -- Kai Schätzl, Berlin, Germany Get your web at Conactive Internet Services: http://www.conactive.com

1 0

No console from VM Manager since 5.1
by Kai Schaetzl 10 Jan '08

10 Jan '08

Since I updated my CentOS 5.0 host machine to 5.1 I cannot open a console from the VM Manager to a guest. Before 5.1 it opened a VM window that looks close to the VMWare window and displays the console in it, so that I can login to the console. Now I get a "pre-window" that shows a "password" textbox, a Login button and a checkbox for "Save this password in your keyring". What is this? What password am I supposed to enter? Entering the root password (of host or guest) does not work. Once I press the Login button it just hangs and when I try to close it I get a warning that it hangs and when I kill it this way the VM console and the VM Manager are destroyed. I can login via "xm console". Kai -- Kai Schätzl, Berlin, Germany Get your web at Conactive Internet Services: http://www.conactive.com

1 0

Problem with CentOS 5.1/xen/RTL8110
by Chris Gow 08 Jan '08

08 Jan '08

Hello: I'm having issues with my CentOS 5.1/Xen installation. If I run the xen-bridge, I seem to get flaky ethernet. By flaky I mean everything seems fine from the host machine, but if I attempt to contact the host machine from another remote machine (eg. my laptop which is on the same subnet as the xen machine, separated by 10 ft of cable and a router) I either get extremely high ping times or Destination Host Unreachable. Ditto with ssh. I've disabled the firewall and it does not make a difference. If I stop the xen-bridge (/etc/xen/scripts/network-bridge stop) then I get what I would consider normal network access to the xen machine. Hardware: Gigabyte GA-G33M-S2H MB (Realtek R8110SC onboard nic) OS: CentOS 5.1 64bit Xen I installed CentOS last night with the Xen kernel, the stock kernel did not support my nic which I was aware of. So I downloaded the r1000 source rpm from the centos wiki, built and installed it. Once I did that the card was detected (an ifconfig would actually display eth0), but it would never get an IP address (the xen machine is configured for dhcp at the moment). Today, I installed the non-xen kernel, applied the non-xen r1000 kernel module and the onboard nic was found and working. I then grabbed the updates from centos, saw that there were some kernel updates, applied the r1000 kernel modules again, restarted and (the non-xen kernel) eth0 was still happy. Good. Restarted again, but booted into the xen kernel, eth0 was still happy. eth0 would get an IP address, and was able to see the outside world. However, the outside world (eg. my laptop) could not see the xen machine or it could inconsistently. That is, ping times would be extremely high (on the order of 2+ seconds) or I would get Destination Host Unreachable errors. Trying to connect via ssh would also be sporadic. Thinking it might be firewall related, I disabled the firewall. There was no change in behaviour. I then disabled the xen-bridge and was able to ping with reasonable numbers (<200ms) and connect via ssh. Just to note though, after I stopped the bridge I immediately tried to connect via ssh/ping and did not get through, so I ran service iptables stop (again) and then was able to get correct network access. I'm not sure if stopping iptables again did anything (I doubt) or I did not leave enough time from stopping the bridge to letting everything get reconfigured. I'm not sure what the problem is or how exactly to troubleshoot it. The NIC is slightly different than the one specified in the CentOS wiki (the wiki mentions RTL8110 and RTL8169SC and mine is a RTL8110SC) but I'm not familiar enough with the devices to know how much of a big deal that is, if any. Also, just to be clear, the problem is other physical machines have a difficult time accessing the xen host pc when the xen-bridge is running. I have not gotten as far as creating a VM yet. Any assistance would be great. thanks -- chris

1 0

VMware images
by Ted Miller 07 Jan '08

07 Jan '08

Can someone please clarify the status of minimal install images of Centos 5 for VMware? I see the work that 'tru' did, but it is still at 5 (not 5.1). Is that the best image available (just run yum update), or what about the minimal image listed on the VMware Virtual Appliance site? Ted Miller Indiana, USA

4 4

tick_divider kernel parameter for guest vm
by Fabian Arrotin 03 Jan '08

03 Jan '08

When upstream released 5.1, everybody wanted to test a new kernel parameter that could adjust the system clock rate at boot time to something else than the standard 1000Hz clock rate. A lot of testings has been done (thanks to Akemi Yagi for her great work) and you can see the results here : http://bugs.centos.org/view.php?id=2189 As you can read at the bottom of the comments, it seems it was a typo in the official RH Release Notes : you'd have to read divider= and *NOT* tick_divider= ! (see http://www.centos.org/docs/5/html/release-notes/as-x86/RELEASE-NOTES-U1-x86…) It seems so to work with the correct kernel parameter and so there is no need to build a kernel-vm for CentOS 5.1 guests .. (it's still needed for example for 4.x ..) Keep on reading the comments on bugs.centos.org for further informations ... I assume that upstream release notes will be corrected to reflect the real parameter -- Fabian Arrotin <fabian.arrotin(a)arrfab.net> Solution ? echo '16i[q]sa[ln0=aln100%Pln100/snlbx]sbA0D4D465452snlbxq' | dc

3 2

No CD/DVD option in virt-manager for para-virtualized VM?
by brad lodgen 03 Jan '08

03 Jan '08

Hey Everyone, I'm fairly new to linux(6 months or so experience) and very new to using Xen virtualization. First post on the mailing list here! Thanks in advance for any and all help attempts, whether it actually helps me or not! I've been reading the Red Hat virtualization guide and in the appendix labs, using virt-manager, it says to: "Specify either CD-ROM or DVD and enter the path to install media. Specify ISO Image location if you will install from an ISO image. Click Forward." Is this option only available in fully-virtualized VM installs? It asks for URL media only when picking para-virtualized VM. Brad

2 1

open-vm-tools for latest CentOS-4 and CentOS-5 kernel-vms
by Johnny Hughes 02 Jan '08

02 Jan '08

All, There are some (very much) testing open-vm-tools RPMS for both the CentOS-4 and CentOS-5 i386 and x86_64 kernel-vm's and standard UP kernels. The kernel-vm's are available here: http://people.centos.org/~tru/kernel-vm/ (The kernel-vm kernels are designed to run inside the VMs at a clock rate of 100HZ instead of 1000Hz.) The open-vm-tools are available here: http://people.centos.org/~hughesjr/open-vm-tools/ The purpose of these RPMS (open-vm-tools) is to replace the VMware-Tools RPMS that come with VMWare. Please remove VMWare-Tools inside the VM if you are going to install these open-vm-tools for testing. There is an xorg.conf file that should be used to replace the one in /etc/X11/ inside the VM Client only (not on your host). The open-vm-tools SHOULD ONLY be installed inside a VM Client and not on the VM Host. These open-vm-tools RPMS are only designed to work with CentOS VM Clients/Guests (should not matter what the host is), though they should also work if RHEL4 or RHEL5 are the VM Clients/Guests. Documentation for the vm tools can be found here: http://open-vm-tools.wiki.sourceforge.net/Packaging and chapter3 here: http://www.vmware.com/pdf/server_vm_manual.pdf (In that PDF file, the installation parts can be ignored, but the configuration parts and usage parts should work) Thanks, Johnny Hughes

5 8

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

2007

virt January 2008