virt September 2008

virt@lists.centos.org

23 participants
10 discussions

amazon web services and CentOS
by Karanbir Singh 28 Sep '08

28 Sep '08

hi, Over the last few months, based on the fact that people were asking me about it, I was in turn asking Amazon what they are doing to make CentOS available on their EC2 platform. After much conversation to / from their end, they are in a situation where they have offered me some access and credits to make the whole CentOS on AWS happen. This is the sort of thing right up the VirtSIG alley, so am wondering if there are people here who would like to help out at all ? -- Karanbir Singh : http://www.karan.org/ : 2522219@icq

4 4

powernow error in Xen VM with new kernel
by Kai Schaetzl 27 Sep '08

27 Sep '08

I just upgraded my first Xen guest to the latest 5.2 kernel (2.6.18- 92.1.13.el5xen) and got a kernel error on reboot: powernow-k8: BIOS error - no PSB or ACPI _PSS objects No problems seen yet. I would think that the guest should not try to do any powernow stuff at all. Has anyone else noticed this? Kai -- Kai Schätzl, Berlin, Germany Get your web at Conactive Internet Services: http://www.conactive.com

1 0

redhat's Qumranet acquisition
by Karanbir Singh 24 Sep '08

24 Sep '08

http://www.redhat.com/about/news/prarchive/2008/qumranet.html I am sure most people would have already seen this, but posting for those who havent already. -- Karanbir Singh : http://www.karan.org/ : 2522219@icq

12 20

how do I access an LV on a xen dom0?
by Tom Diehl 23 Sep '08

23 Sep '08

Hi, I have a machine running several domU's. It has an LV that is currently mounted on dom0, that I would like to access from one of the domU's. If I umount the LV and remove it from the fstab on the dom0, is there a way I can mount it in one of the domU's? If there is a doc that describes how to do this, a pointer would be appreciated. I have tried google but so far no joy. Regards, -- Tom Diehl tdiehl(a)rogueind.com Spamtrap address mtd123(a)rogueind.com

2 2

PCI passthrough problems
by Enderson Maia 22 Sep '08

22 Sep '08

Hello, I'm having a problem, I've seen other people with the same problem, but with USB. Some people solved this with the pciback.permissive option, but in my case it's ignored. I'm using CentOS 5.2 - 2.6.18-92.1.10.el5.centos.plusxen The error return from the console starting the domU is below. I've tried swiotlb=force, pciback.permissive like I said. My configs are: /etc/modprobe.conf ... options pciback hide=(03:01.0) # cat /etc/xen//firewall name = "firewall" uuid = "cd5f6ad8-0682-73fd-3268-fb4a7b85c0ec" maxmem = 256 memory = 256 vcpus = 1 bootloader = "/usr/bin/pygrub" boot = "swiotlb=force" on_poweroff = "destroy" on_reboot = "restart" on_crash = "restart" vfb = [ "type=vnc,vncunused=1" ] disk = [ "phy:/dev/VolGroup00/firewall.disk,xvda,w" ] vif = [ "mac=00:16:3e:0e:53:5a,bridge=virbr0" ] pci = [ '0000:03:01.0' ] # rpm -qa | grep xen xen-libs-3.0.3-64.el5_2.1 xen-3.0.3-64.el5_2.1 kernel-xen-2.6.18-92.el5 kernel-xen-2.6.18-92.1.10.el5.centos.plus --- Iniciando interface eth0: [ OK ] Iniciando interface eth1: Determinando informação IP para o eth1...Fatal DMA error! Please use 'swiotlb=force' ------------[ cut here ]------------ kernel BUG at arch/i386/kernel/pci-dma-xen.c:401! invalid opcode: 0000 [#1] SMP last sysfs file: /class/net/eth1/address Modules linked in: xennet ip_conntrack_netbios_ns ipt_REJECT xt_state ip_conntrack nfnetlink iptable_filter ip_tables ip6t_REJECT xt_tcpudp ip6table_filter ip6_tables x_tables ipv6 xfrm_nalgo crypto_api dm_mirror dm_multipath dm_mod parport_pc lp parport 3c59x pcspkr mii xenblk ext3 jbd uhci_hcd ohci_hcd ehci_hcd CPU: 0 EIP: 0061:[<c040a39c>] Not tainted VLI EFLAGS: 00010292 (2.6.18-92.el5xen #1) EIP is at dma_map_single+0x12d/0x147 eax: 0000002f ebx: 00000000 ecx: 00000001 edx: f5416000 esi: 38f59e02 edi: 00000001 ebp: 0000004e esp: c071bdbc ds: 007b es: 007b ss: 0069 Process swapper (pid: 0, ti=c071b000 task=c066e2c0 task.ti=c06e6000) Stack: c0621abd c07fc448 008a8e02 00000000 cf914400 cf5db200 c07fc448 00000140 d1232e6e 00000001 c08a8e02 cf914000 cf88c200 00000002 cf5db200 0000004e d126c000 cf914000 0000004e c08a8e02 cf88c140 cf88c140 cf88c200 c07c4c88 Call Trace: [<d1232e6e>] boomerang_start_xmit+0x141/0x352 [3c59x] [<c05aec59>] dev_hard_start_xmit+0x198/0x1ee [<c05bcaf6>] __qdisc_run+0xe4/0x19a [<c05b0544>] dev_queue_xmit+0x1e0/0x2e8 [<c05b4b24>] neigh_resolve_output+0x1b2/0x1dc [<d136eb2e>] ip6_output2+0x256/0x2a7 [ipv6] [<d136f656>] ip6_output+0x9c7/0x9f3 [ipv6] [<d137c8c0>] dst_output+0x0/0x7 [ipv6] [<c05c3d12>] nf_hook_slow+0x3a/0x90 [<d137c8c0>] dst_output+0x0/0x7 [ipv6] [<d137e4d6>] ndisc_send_ns+0x3a8/0x47a [ipv6] [<c06090d8>] _spin_lock_irqsave+0x8/0x28 [<c042914a>] lock_timer_base+0x15/0x2f [<d1375ddb>] addrconf_dad_timer+0xc3/0xdf [ipv6] [<c0434244>] hrtimer_run_queues+0x14b/0x163 [<d1375d18>] addrconf_dad_timer+0x0/0xdf [ipv6] [<c0429600>] run_timer_softirq+0x101/0x15c [<c042613e>] __do_softirq+0x5e/0xc3 [<c0406edf>] do_softirq+0x56/0xaf [<c0406e80>] do_IRQ+0xa5/0xae [<c0549913>] evtchn_do_upcall+0x64/0x9b [<c04055d9>] hypervisor_callback+0x3d/0x48 [<c0408632>] raw_safe_halt+0x8c/0xaf [<c040321a>] xen_idle+0x22/0x2e [<c0403339>] cpu_idle+0x91/0xab [<c06eb9f5>] start_kernel+0x37a/0x381 ======================= Code: c0 74 07 8b 08 8b 58 04 eb 05 83 c9 ff 31 db 89 da 89 c8 f7 d2 f7 d0 21 fa 21 f0 89 d1 09 c1 74 13 68 bd 1a 62 c0 e8 de 7b 01 00 <0f> 0b 91 01 9e 1a 62 c0 5d f0 83 04 24 00 83 c4 0c 89 f0 89 fa EIP: [<c040a39c>] dma_map_single+0x12d/0x147 SS:ESP 0069:c071bdbc <0>Kernel panic - not syncing: Fatal exception in interrupt --- If you need more information, just tell -- Enderson Maia

1 0

Status of kernel divider option in 2.6.9-78.0.1
by Jeff 20 Sep '08

20 Sep '08

Under 4.6, we recompiled the kernel with HZ=100 for improved time-keeping in VMware guests. I've read about the backporting of the divider patch into RHEL/CentOS 4.7, but it sounds like it also comes with some bugs. I have been unable to determine the current status of the divider option in the latest 4.7 kernel update. I have experimented with "divider=10" and it works with no major problems, except that the guest clock drifts ahead very slightly. So far, it looks like several seconds of drift per day. Not a show-stopper, but still a concern. With HZ=100, our time sync was spot on. In both cases we are using "clock=pit nosmp noapic" and tools.syncTime="TRUE". If it matters, our host machines are CentOS 4.6 x86_64. Can anyone comment on the current status of bugs with the divider option? Is anyone getting accurate time sync with divider=10 with 2.6.9-78.0.1? Is there anything else to watch out for? Is there any fix for the slight time drift with divider=10? I am primarily concerned with i686. Currently, the SRPM for -78.0.1 is missing from the mirrors, so if I want to update, I have to rely on the divider option for my VMware guests. It also appears that Tru has not yet added any -78 RPMs to his kernel-vm project. -- Jeff

2 4

vmware on centos cluster
by Mattias Hemmingsson 11 Sep '08

11 Sep '08

Hello I have one question about cluster and vmware. I have about 10 computers they are all old once from 1g and 256 ram. And now im thinking of putting them all toghter in one cluster. This cluster should be an high performing so all computers share they performance. Now on this cluster i whould like to install one vmware server and make that vmware server get the power from all these computers. And install a os in that vmware server. Is this the way passible to do ? // matte

2 1

Xen kickstart and virt-manager
by Brett Worth 11 Sep '08

11 Sep '08

Hi again. I'm pleased to say that my experimentation with ocfs2 has been very successful and I'm considering doing some real work with it. My current question is related to virt-manager and the use of kickstart to install paravirtualised domU's. I'm happily doing fully-virtualised installs with this tool but due to poor network performance I thought it time to do some para-virtualised testing. In the dialog where I'm asked to supply the URL for the install repository and the kickstart location I'm unable to use the "file://..." construct so I'm left with a selection that only includes locations on the network. Now my network interface information for the domU is in the kickstart file so when I move forward in the tool there's a message saying that its trying to determine a network address, presumably via DHCP (I am unable to put a dhcp server on this network). From the command line I can use wget against the kickstart file at the URL I'm providing so that's OK. I'm obviously overlooking something but its currently got me puzzled. So how is this supposed to work? Am I supposed to do the network configuration for the installation somewhere else? Brett

3 4

Disappearing TAP interface
by James Roman 08 Sep '08

08 Sep '08

I have a CentOS 5.2 Xen host, running several guest domains. Each guest domain has a single interface assigned to it. I am using both the network-bridge and vif-bridge scripts. On my only windows guest, the tap interface connected to the bridge keeps getting disabled. If I check "brctl show" the tap0 interface is not present, and there will be several message log entries indicating that the interface tap0 is being disabled. This means that the guest becomes invisible. I can not open a VNC console or connect to it from the network. From the hosts log files, it appears that the OS keeps running, even though it it is unreachable. (this is a little hard to guage, because the times in the eventlog keep jumping around.) The only way I have been able to restore a connection is to destroy the domain and re-create it. I can not find much in the various Xen logs that identifies what triggers the tap interface to be disabled. I have several questions. Is there any way to manually rebuild the tap interface to get access to the host again (either through VNC or network)? Maybe through the virsh command or calling the vif-bridge or qemu-ifup command directly? Is there another way that I could use to maintain a connection to the guest? Does anyone have any idea what causes the tap interface to go down? --qemu-dm log---------------------------------------------------- domid: 1 qemu: the number of cpus is 2 *** image /local/domain/0/backend/vbd/1/768 format raw driver 0x80e7940 *** image /local/domain/0/backend/vbd/1/5632 format raw driver 0x80e7940 *** image /local/domain/0/backend/vbd/1/832 format raw driver 0x80e7940 Watching /local/domain/1/logdirty/next-active xs_read(): vncpasswd get error. /vm/e2a537ca-5ef5-e87d-6505-855529af3ea4/vncpasswd. char device redirected to /dev/pts/1 qemu_map_cache_init nr_buckets = 4000 shared page at pfn 3ffff buffered io page at pfn 3fffd xs_read(/vm/e2a537ca-5ef5-e87d-6505-855529af3ea4/rtc/timeoffset): read error I/O request not ready: 0, ptr: 0, port: 0, data: 0, count: 0, size: 0 I/O request not ready: 0, ptr: 0, port: 0, data: 0, count: 0, size: 0 Triggered log-dirty buffer switch xs_write(/vm/e2a537ca-5ef5-e87d-6505-855529af3ea4/rtc/timeoffset, rtc/timeoffset): write error xs_write(/vm/e2a537ca-5ef5-e87d-6505-855529af3ea4/rtc/timeoffset, rtc/timeoffset): write error --------------------------------------------------------------------------- name = "vm-002_win" uuid = "e2a537ca5ef5e87d6505855529af3ea4" maxmem = 2048 memory = 1024 vcpus = 2 builder = "hvm" kernel = "/usr/lib/xen/boot/hvmloader" boot = "c" pae = 1 acpi = 1 apic = 1 on_poweroff = "destroy" on_reboot = "restart" on_crash = "restart" device_model = "/usr/lib/xen/bin/qemu-dm" sdl = 0 vnc = 1 vncunused = 1 keymap = "en-us" disk = [ "file:/var/lib/xen/images/vm-002_win.img,hda,w", "file:/data/os-install-media/Win2003plusSP1.iso,hdc:cdrom,r", "phy:/dev/VolGroup00/LVssavmWin001Data,hdb,w" ] vif = [ "ip=192.168.2.103,mac=00:16:3e:0d:7f:be,bridge=xenbr0,type=ioemu" ] serial = "pty" -- James D. Roman IT Network Administration Terranet Inc.On contract to: Science Systems and Applications, Inc.

1 0

relocation, migration
by Paras pradhan 02 Sep '08

02 Sep '08

Hi, I am running a cluster having 2 nodes using red hat cluster suite. node 1 has a para virtualized guest(centOS) running under Xen. My question is when node1 is rebooted, guest is automatically relocated to node 2 . Instead of relocation, is migration possible in this case which can result in Zero down time? Thanks in adv Paras.

3 3

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

2007

virt September 2008