May 2008 - xen - Fedora mailing-lists

Guest shutdown problems
by Christian Axelsson 02 Apr '09

02 Apr '09

Hello I have a problem shutting down xen guests. Using xm shutdown guests get shut down but gets stuck in state '---s-d' or sometimes '------'. When trying to clone a domain when in this state (my original purpose of the whole operation) I'll get the error: [root@hydra virtinst--devel]# ./virt-clone -o minimal -n new_img -f /var/lib/xen/images/new_img.img ERROR: virDomainGetXMLDesc() failed failed Xen syscall xenDaemonDomainDumpXMLByID failed to find this domain -490299505 The same errors occurs when for example trying to attach to the console using virsh. I have tried to use 'xm destroy' to kill the guest the hard way but it has no effect - the state remains unchanged. I have also tried this on a few different guest installations with the same result. A thing worth noting is that the output from 'xm list --long' differs, I've attached the out put pre boot, after boot and after shutdown. Note how all the devices in the guests are missing after shutdown. Both the hosts and the guests are fedora 8 installations. Regards, Christian Axelsson smiler(a)lanil.mine.nu [?1034h(domain (domid 0) (on_crash restart) (uuid 00000000-0000-0000-0000-000000000000) (bootloader_args ) (vcpus 2) (name Domain-0) (on_poweroff destroy) (on_reboot restart) (bootloader ) (maxmem 16777215) (memory 1491) (shadow_memory 0) (cpu_weight 256) (cpu_cap 0) (features ) (on_xend_start ignore) (on_xend_stop ignore) (cpu_time 1644.84369405) (online_vcpus 2) (image (linux (kernel ))) (status 2) (state r-----) ) (domain (domid 2) (on_crash restart) (uuid a7638797-e237-3891-5e64-390f828238ca) (bootloader_args ) (vcpus 1) (name minimal) (on_poweroff destroy) (on_reboot restart) (bootloader /usr/bin/pygrub) (maxmem 512) (memory 512) (shadow_memory 0) (cpu_weight 256) (cpu_cap 0) (features ) (on_xend_start ignore) (on_xend_stop ignore) (start_time 1206360333.14) (cpu_time 9.753408915) (online_vcpus 1) (image (linux (kernel ) (notes (FEATURES 'writable_page_tables|writable_descriptor_tables|auto_translated_physmap|pae_pgdir_above_4gb|supervisor_mode_kernel' ) (VIRT_BASE 18446744071562067968) (GUEST_VERSION 2.6) (PADDR_OFFSET 18446744071562067968) (GUEST_OS linux) (HYPERCALL_PAGE 18446744071564189696) (LOADER generic) (SUSPEND_CANCEL 1) (ENTRY 18446744071564165120) (XEN_VERSION xen-3.0) ) ) ) (status 2) (state -b----) (store_mfn 196619) (console_mfn 196618) (device (vif (bridge xenbr0) (mac 00:16:3e:3f:93:b8) (script vif-bridge) (uuid 94afd732-920b-2e0b-b3d5-e79174754a80) (backend 0) ) ) (device (vbd (uname file:/var/lib/xen/images/minimal.img) (uuid 8f4f4da3-5f8a-3fee-28e8-41dc49e876cd) (mode w) (dev xvda:disk) (backend 0) (bootable 1) ) ) (device (console (protocol vt100) (location 2) (uuid 0046f2d3-058b-d524-9273-f1dac2ca950b) ) ) ) [?1034h(domain (domid 0) (on_crash restart) (uuid 00000000-0000-0000-0000-000000000000) (bootloader_args ) (vcpus 2) (name Domain-0) (on_poweroff destroy) (on_reboot restart) (bootloader ) (maxmem 16777215) (memory 1491) (shadow_memory 0) (cpu_weight 256) (cpu_cap 0) (features ) (on_xend_start ignore) (on_xend_stop ignore) (cpu_time 1648.92600832) (online_vcpus 2) (image (linux (kernel ))) (status 2) (state r-----) ) (domain (domid 2) (on_crash restart) (uuid a7638797-e237-3891-5e64-390f828238ca) (bootloader_args ) (vcpus 1) (name minimal) (on_poweroff destroy) (on_reboot restart) (bootloader /usr/bin/pygrub) (maxmem 512) (memory 512) (shadow_memory 0) (cpu_weight 256) (cpu_cap 0) (features ) (on_xend_start ignore) (on_xend_stop ignore) (start_time 1206360333.14) (cpu_time 13.048743365) (online_vcpus 1) (image (linux (kernel ) (notes (FEATURES 'writable_page_tables|writable_descriptor_tables|auto_translated_physmap|pae_pgdir_above_4gb|supervisor_mode_kernel' ) (VIRT_BASE 18446744071562067968) (GUEST_VERSION 2.6) (PADDR_OFFSET 18446744071562067968) (GUEST_OS linux) (HYPERCALL_PAGE 18446744071564189696) (LOADER generic) (SUSPEND_CANCEL 1) (ENTRY 18446744071564165120) (XEN_VERSION xen-3.0) ) ) ) (status 0) (state ---s-d) (store_mfn 196619) (console_mfn 196618) ) [?1034h(domain (domid 0) (on_crash restart) (uuid 00000000-0000-0000-0000-000000000000) (bootloader_args ) (vcpus 2) (name Domain-0) (on_poweroff destroy) (on_reboot restart) (bootloader ) (maxmem 16777215) (memory 1491) (shadow_memory 0) (cpu_weight 256) (cpu_cap 0) (features ) (on_xend_start ignore) (on_xend_stop ignore) (cpu_time 1635.21430615) (online_vcpus 2) (image (linux (kernel ))) (status 2) (state r-----) ) (domain (on_crash restart) (uuid a7638797-e237-3891-5e64-390f828238ca) (bootloader_args ) (vcpus 1) (name minimal) (on_poweroff destroy) (on_reboot restart) (bootloader /usr/bin/pygrub) (maxmem 512) (memory 512) (shadow_memory 0) (cpu_weight 256) (cpu_cap 0) (features ) (on_xend_start ignore) (on_xend_stop ignore) (start_time 1206309092.82) (cpu_time 0.0) (image (linux (kernel ) (notes (FEATURES 'writable_page_tables|writable_descriptor_tables|auto_translated_physmap|pae_pgdir_above_4gb|supervisor_mode_kernel' ) (VIRT_BASE 18446744071562067968) (GUEST_VERSION 2.6) (PADDR_OFFSET 18446744071562067968) (GUEST_OS linux) (HYPERCALL_PAGE 18446744071564189696) (LOADER generic) (SUSPEND_CANCEL 1) (ENTRY 18446744071564165120) (XEN_VERSION xen-3.0) ) ) ) (status 0) (device (vif (bridge xenbr0) (mac 00:16:3e:3f:93:b8) (backend 0) (uuid 94afd732-920b-2e0b-b3d5-e79174754a80) (script vif-bridge) ) ) (device (vbd (uuid 8f4f4da3-5f8a-3fee-28e8-41dc49e876cd) (bootable 1) (driver paravirtualised) (dev xvda:disk) (uname file:/var/lib/xen/images/minimal.img) (mode w) (backend 0) ) ) (device (console (protocol vt100) (location 2) (uuid 0046f2d3-058b-d524-9273-f1dac2ca950b) ) ) )

3 3

Cannot start HVM DomU
by Robert Locke 04 Jul '08

04 Jul '08

This has newly started, so I am presuming that it is perhaps related to some recently updated package on F8 in say the last week or so. I have a WindowsXP DomU that when I try to start from virt-manager gives the following error: Error starting domain: virDomainCreate() failed POST operation failed: (xend.err 'int argument required') Details shows: Traceback (most recent call last): File "/usr/share/virt-manager/virtManager/engine.py", line 472, in run_domain vm.startup() File "/usr/share/virt-manager/virtManager/domain.py", line 379, in startup self.vm.create() File "/usr/lib/python2.5/site-packages/libvirt.py", line 240, in create if ret == -1: raise libvirtError ('virDomainCreate() failed', dom=self) libvirtError: virDomainCreate() failed POST operation failed: (xend.err 'int argument required') I have no idea what other information to supply or where to particularly find it since this has generally "just worked" for the occasional running of a Windows instance that I have needed for the last several months. Thanks for any help, --Rob

2 2

RE: [Fedora-xen] Strange mouse issue (F7 & F8 xen kernel with vncviewer running)
by Dustin Henning 30 May '08

30 May '08

I spoke too soon... I just tried krdc and the same thing happened before I could even try to connect to anything with it (or even determine if it supports vnc, I don't remember). This is a quad core, and I have, in all instances, confirmed that each of the four vcpus in dom0 is tied to a different core. Additionally, I have never had more than one DomU per core on top of Dom0 using each core. Finally, in this latest instance, only one DomU was doing anything, and it could not have been using but 100% of one core, so the other three were, for lack of a better word, bored. -----Original Message----- From: fedora-xen-bounces(a)redhat.com [mailto:fedora-xen-bounces@redhat.com] On Behalf Of Dustin Henning Sent: Friday, May 30, 2008 16:48 To: fedora-xen(a)redhat.com Subject: [Fedora-xen] Strange mouse issue (F7 & F8 xen kernel with vncviewer running) I have been experiencing this (rather strange) issue for as long as I have been a member of this group, and I am curious as to whether anyone else has experienced it. The mouse will continue on in the direction I am moving it after I stop, change direction, or even suddenly reverse direction. Once this happens, the mouse inevitably ends up in one of the four corners of the screen, and will eventually stop trying to travel in that direction, at which point I can control it again. It seems like hitting some keys may cause this to happen sooner. The keys I most often go for are Esc, Alt, Ctrl, Super, Alt+F1, Arrows. Which keys usually depends on what is up in the vncviewer window, where I don't want to issue the wrong command to whatever client I may be working with. I have made the following observations regarding this behavior, though I can't say with absolute certainty that they are all 100% accurate and/or consistent: -keyboard and mouse are PS/2 -in all instances, hvm guests were running, consuming as much as 5 out of 8 GiB of memory on this machine -never experienced in standard kernel on F7 or F8 (even with multiple kvm guests and vncviewers running, though not to consume 5 of 8 GiB) -originally, in F7, vncviewer may not have been running every time it happened -now, in F8, vncviewer has definitely been running every time it happened, but not necessarily connected to anything -same issue was previously experienced on another machine with less memory (and less guests) on F7 Is this a known issue of any sort? It has recently occurred to me that I should try a different client, and I will do that at some point, but I figured it wouldn't hurt to see if there have been any other similar experiences out there in the meantime. Thanks, Dustin -- Fedora-xen mailing list Fedora-xen(a)redhat.com https://www.redhat.com/mailman/listinfo/fedora-xen

1 0

Strange mouse issue (F7 & F8 xen kernel with vncviewer running)
by Dustin Henning 30 May '08

30 May '08

I have been experiencing this (rather strange) issue for as long as I have been a member of this group, and I am curious as to whether anyone else has experienced it. The mouse will continue on in the direction I am moving it after I stop, change direction, or even suddenly reverse direction. Once this happens, the mouse inevitably ends up in one of the four corners of the screen, and will eventually stop trying to travel in that direction, at which point I can control it again. It seems like hitting some keys may cause this to happen sooner. The keys I most often go for are Esc, Alt, Ctrl, Super, Alt+F1, Arrows. Which keys usually depends on what is up in the vncviewer window, where I don't want to issue the wrong command to whatever client I may be working with. I have made the following observations regarding this behavior, though I can't say with absolute certainty that they are all 100% accurate and/or consistent: -keyboard and mouse are PS/2 -in all instances, hvm guests were running, consuming as much as 5 out of 8 GiB of memory on this machine -never experienced in standard kernel on F7 or F8 (even with multiple kvm guests and vncviewers running, though not to consume 5 of 8 GiB) -originally, in F7, vncviewer may not have been running every time it happened -now, in F8, vncviewer has definitely been running every time it happened, but not necessarily connected to anything -same issue was previously experienced on another machine with less memory (and less guests) on F7 Is this a known issue of any sort? It has recently occurred to me that I should try a different client, and I will do that at some point, but I figured it wouldn't hurt to see if there have been any other similar experiences out there in the meantime. Thanks, Dustin

1 1

Xen Documentation
by Gastón Keller 30 May '08

30 May '08

Hello, everybody. I have been struggling with Xen's documentation lately, not finding where to look for accurate and up-to-date info. I remember having read in this list months ago someone saying that the documentation was quite outdated. For example, the xm command _delete_ does not appear in the man pages and many guides still do reference to the guest's config files being stored at /etc/xen (two days ago I discovered that now Xen works in a different way). My objective is not only to learn how to use Xen, but understand how it works. I'm doing my best looking for info in the net and going through docs in the Xen Wiki, but I might be missing something in my search. Therefore, I would really appreciate if someone could point me to some documentation I should read, or whatever else you could think of. Thanks, Gaston -- La única verdad es la realidad.

6 11

Re: [Fedora-xen] Xen Documentation
by Emre Erenoglu 30 May '08

30 May '08

I totally share your view. :) I'm in the same condition. Emre On Fri, May 30, 2008 at 2:12 PM, Dustin Henning <Dustin.Henning(a)prd-inc.com> wrote: > Emre, > As you may already know, I still use the config files in /etc/xen > and don't mess with the virsh stuff. It all certainly worked fine in F7, > and though I haven't fully tested, it appears that it all works fine in F8 > as well. I use them for one, because I learned with them, and for two, > because in the little time I have spent messing with the gui tools, I > haven't found near as much control. That isn't to say that more can be done > with the old way, but instead to say that I can do more with the old way. I > don't know if there are any plans to deprecate that stuff, but I hope that > if there ever are, news makes the list before I end up on a version where I > have no clue how to do anything! > Dustin > > From: fedora-xen-bounces(a)redhat.com [mailto:fedora-xen-bounces@redhat.com] > On Behalf Of Emre ERENOGLU > Sent: Friday, May 30, 2008 04:30 > To: Gastón Keller > Cc: fedora-xen(a)redhat.com > Subject: Re: [Fedora-xen] Xen Documentation > > Hmm, > On Fri, May 30, 2008 at 2:06 AM, Gastón Keller <gastonkeller(a)gmail.com> > wrote: > On Thu, May 29, 2008 at 7:53 PM, Emre ERENOGLU <erenoglu(a)gmail.com> wrote: > > Hi Gaston, > > > > It's true that some documentation is outdated. However, for the > "/etc/xen" > > isssue, my custom Xen 3.2 setup still stores ( and I do store there also) > my > > config files there. > > > > Maybe what you're saying is "Fedora Specific" which I could understand as > > it's fedora mailing list. > _The Fedora team has followed the Xensource model and begun to store > all VM configuration details in a database, referred to as xenstore._ > Source: > http://enterpriselinuxlog.blogs.techtarget.com/2007/06/07/fedora-7-xen-firs… > > OK, I didn't know that. However, I find this xenstore thing a bit > complicated. > > Then, you might be right. I interpreted it was a modification from > Xen, not from Fedora (previously introduced by Xensource). > > So, what is the advantage behind this movement? > > I think it's managegability. > > > Thanks, > Gaston > > However, I never found the flexibility and control > > in the Fedora way of Xen implementation (all those xml stuff etc.). most > > probably, I'm wrong :) > > > > Emre > > > > On Thu, May 29, 2008 at 9:22 PM, Gastón Keller <gastonkeller(a)gmail.com> > > wrote: > >> > >> Hello, everybody. I have been struggling with Xen's documentation > >> lately, not finding where to look for accurate and up-to-date info. I > >> remember having read in this list months ago someone saying that the > >> documentation was quite outdated. For example, the xm command _delete_ > >> does not appear in the man pages and many guides still do reference to > >> the guest's config files being stored at /etc/xen (two days ago I > >> discovered that now Xen works in a different way). > >> > >> My objective is not only to learn how to use Xen, but understand how > >> it works. I'm doing my best looking for info in the net and going > >> through docs in the Xen Wiki, but I might be missing something in my > >> search. Therefore, I would really appreciate if someone could point me > >> to some documentation I should read, or whatever else you could think > >> of. > >> > >> Thanks, > >> Gaston > >> > >> -- > >> La única verdad es la realidad. > >> > >> -- > >> Fedora-xen mailing list > >> Fedora-xen(a)redhat.com > >> https://www.redhat.com/mailman/listinfo/fedora-xen > > > > > > > > -- > > Emre > > > -- > La única verdad es la realidad. > > > > -- > Emre > > > -- Emre

1 0

RE: [Fedora-xen] HVM DomU on F9?
by Dustin Henning 30 May '08

30 May '08

For anyone who might have been following this thread, I accidentally forgot to add fedora-xen to the recipient list before sending my last reply. It is below, along with a response from Avi. I updated my F8 box, and sure enough, kvm-60 was installed. I thought I had just done that and kvm-55 was where it stopped, but I must be crazy or something. Also, I got the gplpv drivers by James Harper installed and functional as of version 0.9.2 (I had no luck with 0.9.0 or any prior version). That said, I probably won't go any further with my quest for kvm, at least not in the near future. The stdvga emulation kvm uses (or at least used on kvm-47) was nice with the standard vesa drivers referenced on the kvm site (went higher than 1024x768), and those drivers don't work for the stdvga=1 option in xen, but I think the potential performance gain from the multiple paravirtual drivers will be more important for me. Also, as I didn't follow up on it, I did try to install CentOS 5.1 on this machine and it does not support the network card, so I am on Fedora 8 as originally recommended by Emre. Dustin -----Original Message----- From: Avi Kivity [mailto:avi@qumranet.com] Sent: Friday, May 30, 2008 03:13 To: Dustin.Henning(a)prd-inc.com Subject: Re: [Fedora-xen] HVM DomU on F9? [please use the list for questions like this] Dustin Henning wrote: > It doesn't look like F8 is likely to get up to kvm-58 either. F8 is up to kvm-60. > I > prefer to stick with the Fedora repositories for the machine I am working > with, as I don't want to run into compatibility issues on account of some > future update. However, I want to try to use the network driver, as I am > hopeful that it might make an apache application I am running perform > better. It is also possible that I really need disk drivers for that. > Would compiling a newer version of kvm on Fedora 8 be risky (in terms of > interoperability problems after some future update)? No. > Also, assuming the > answer to that is no, do I need to remove any certain packages first to make > sure they don't update and overwrite it? No. Do a ./configure --prefix=/opt/kvm-69 and all files will be placed there. > Finally, on the Fedora repo > versions of KVM for F7 and F8, I can't create a VM with 2 GiB of RAM. Was > this a known kvm bug at some point (or worse, is it still?), or could it be > a problem with Fedora (I can do this fine in xen)? Thanks, > For more than 2GB of RAM in a guest, you need a 64-bit hypervisor. Large guest support was added in kvm-47 (2TB guests were booted). -- Do not meddle in the internals of kernels, for they are subtle and quick to panic.

1 0

Is pci-pasthrough enabled for F9 DomU?
by snowcrash+xen＠gmail.com 29 May '08

29 May '08

i've F8 as Dom0. for F8 DomU, pci-passthru devices are visible via lspci *in* DomU. for a new install of F9 DomU, lcpci is (empty). there was discussion that PCI-PassThrough might not make it into F9, http://lists.xensource.com/archives/html/xen-devel/2007-12/msg00454.html "> F9 may well end up lacking features compared to Xen in Fedora 8 & earlier (eg no > PCI device passthrough, or CPU hotplug)" but release notes, http://fedoraproject.org/wiki/Features/XenPvops#head-e15c9305e8d715da9d904a… make no mention one way or other. *is* pci-passthru known/supposed to work in F9 DomU? if so, then i've got a different issue ...

7 22

configure: error: GRUB requires a working absolute objcopy; upgrade your binutils
by Dustin Henning 28 May '08

28 May '08

I hope this isn't off topic since the bug I am trying to resolve only affects the xen kernel... I am trying to compile grub on Fedora 8 because I need to apply the patch for bug 250299. First, I downloaded and installed http://mirrors.kernel.org/fedora/releases/8/Fedora/source/SRPMS/grub-0.97-1… Next, I extracted the included grub-0.97.tar.gz to /usr/src/ At this point, if I run configure, I will get the error mentioned in the subject, as follows: [CODE]checking for a BSD-compatible install... /usr/bin/install -c checking whether build environment is sane... yes checking for gawk... gawk checking whether make sets $(MAKE)... yes checking build system type... x86_64-unknown-linux-gnu checking host system type... x86_64-unknown-linux-gnu checking whether to enable maintainer-specific portions of Makefiles... no checking for gcc... gcc checking for gcc... (cached) gcc checking for C compiler default output file name... a.out checking whether the C compiler works... yes checking whether we are cross compiling... no checking for suffix of executables... checking for suffix of object files... o checking whether we are using the GNU C compiler... yes checking whether gcc accepts -g... yes checking for gcc option to accept ANSI C... none needed checking for style of include used by make... GNU checking dependency style of gcc... gcc3 checking dependency style of gcc... (cached) gcc3 checking for ranlib... ranlib checking whether optimization for size works... yes checking whether gcc has -fno-stack-protector... yes checking whether -Wundef works... yes checking whether -falign-loops works... yes checking for objcopy... objcopy checking if C symbols get an underscore after compilation... no checking whether objcopy works for absolute addresses... no configure: error: GRUB requires a working absolute objcopy; upgrade your binutils[/CODE] However, I am running binutils 2.17.50.0.18 which is the latest in the fedora repo, so I cannot upgrade, and I do not know how to downgrade. I saw another issue where this same error was happening, and someone suggested downloading ftp://alpha.gnu.org/gnu/grub/grub-0.97.tar.gz and using it. I tried this and determined that that is the exact file included in the srpm I had downloaded, and the error still surfaces when I try it. Additionally, if I apply the included grub-fedora-8.patch, the configure file disappears. I can get it from the tar again, but I will still get the same error, so while I need to know how I am supposed to compile after applying that patch (which I apply before the patch for the aforementioned bug), but that point is moot until I can determine how to fix this binutils issue.

1 1

I am now on Refriendz!
by Sameer Gupta 28 May '08

28 May '08

Hi! I would like to invite you to visit my Refriendz page and see my latest photos. In order to visit my space, you must go to: http://www.refriendz.com/?do=Login.Invite&rid=sam9823444029&email=fedora-xe… (If this link does not work, please copy and paste it into your browser or go to www.refriendz.com and enter 'sam9823444029' as Invitation ID to Login to the web site.) P.S. Refriendz is Invitation-Only, so do not miss your chance to visit my page! Cheers! Sameer ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Unsubscribe: to opt out of ALL future emails from Refriendz, visit: http://www.refriendz.com/?do=Login.RemoveEntry&email=fedora-xen@redhat.com Please do not reply directly to this email. This mailbox is not monitored and you will not receive a response. Refriendz Limited, PO BOX 1184, Luton, Bedfordshire, LU1 9AT, UK.

1 0

Cannot erase unsuccessfully created guests
by Gastón Keller 27 May '08

27 May '08

Hello, list. I've been playing with Xen lately and it was only until yesterday that I could successfully create a guest with Fedora 8 installed on it (the one with name _cero_). The problem is that all the previous attempts have generated _records_ that I cannot erase. The following is from the terminal: [root@alfa11 gkeller2]# virsh list Id Name State ---------------------------------- 0 Domain-0 running 10 cero blocked [root@alfa11 gkeller2]# /usr/sbin/xm list Name ID Mem VCPUs State Time(s) Domain-0 0 1237 2 r----- 3085.3 cero 10 256 1 -b---- 728.5 fedora0 256 1 0.0 guest0 256 1 0.0 vm0 512 1 0.0 xen0 512 1 0.0 zero 256 1 0.0 [root@alfa11 gkeller2]# virsh destroy xen0 error: Failed to destroy domain xen0 [root@alfa11 gkeller2]# /usr/sbin/xm destroy xen0 [root@alfa11 gkeller2]# /usr/sbin/xm list Name ID Mem VCPUs State Time(s) Domain-0 0 1237 2 r----- 3086.2 cero 10 256 1 -b---- 728.6 fedora0 256 1 0.0 guest0 256 1 0.0 vm0 512 1 0.0 xen0 512 1 0.0 zero 256 1 0.0 [root@alfa11 gkeller2]# It seems virsh doesn't detect those records, but xm does. Also, virsh cannot destroy them (what makes sense if it cannot see them), neither can xm. My problem is that, as far as I know, those _records_ of failed guest have memory assigned from dom0 and they affect the creation of new guests. Any suggestions? Thank you in advance, Gaston -- La única verdad es la realidad.

2 2

fedora 9 guest kernel panic
by Andriy Gapon 26 May '08

26 May '08

We have FC6 Dom0 and recently upgraded one of DomUs to Fedora 9. After that we started to get seemingly random panics for that DomU. Below is some info about that. Any ideas of what might be wrong? Any suggestions on fixing? Kernel BUG at ffffffff80465fc0 [verbose debug info unavailable] invalid opcode: 0000 [1] CPU 0 Modules linked in: nfs nfsd lockd nfs_acl auth_rpcgss exportfs autofs4 fuse sunrpc ipv6 dm_multipath pcspkr xen_netfront dm_snapshot dm_zero dm_mirror dm_mod xen_blkfront ext3 jbd mbcache uhci_hcd ohci_hcd ehci_hcd Pid: 1353, comm: automount Not tainted 2.6.25-2.fc9.x86_64.xen #1 RIP: e030:[<ffffffff80465fc0>] [<ffffffff80465fc0>] xen_failsafe_callback+0x0/0x10 RSP: e02b:ffff880007375e00 EFLAGS: 00010012 RAX: 000000000000000d RBX: ffff88000a9c2000 RCX: ffffffff80465fc0 RDX: ffffffff80627210 RSI: 0000000000000001 RDI: ffffffff80627350 RBP: ffff880007375ea0 R08: ffffffff8063d730 R09: 0000000000000000 R10: 0000003401024b21 R11: 0000000000000212 R12: ffff88001f442000 R13: ffff88000a9c2470 R14: 0000000000000000 R15: ffff88001f429c78 FS: 0000000000000000(0000) GS:ffffffff805bf000(0000) knlGS:0000000000000000 CS: e033 DS: 0000 ES: 0000 CR0: 000000008005003b CR2: 00007f9409bd1780 CR3: 000000001dd9e000 CR4: 0000000000000660 DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000 DR3: 0000000000000000 DR6: 00000000ffff0ff0 DR7: 0000000000000000 Process automount (pid: 1353, threadinfo ffff88001f428000, task ffff88001f442000) Stack: ffffffff806171aa 0000000000000212 0000000000000000 0000000000000000 0000000000000063 0000000000000000 ffffffff806171a8 000000010000e030 0000000000000012 ffff880007375e58 000000000000e02b 0000000000000000 Call Trace: Code: 0f 07 66 0f 1f 84 00 00 00 00 00 48 8b 0c 24 4c 8b 5c 24 08 48 83 c4 10 6a 00 50 48 8d 05 19 00 00 00 e9 54 fb ff ff 0f 1f 40 00 <0f> 0b 66 66 66 66 66 2e 0f 1f 84 00 00 00 00 00 48 89 fc 65 ff RIP [<ffffffff80465fc0>] xen_failsafe_callback+0x0/0x10 RSP <ffff880007375e00> ---[ end trace ffae49894bdd5d90 ]--- BUG: soft lockup - CPU#0 stuck for 61s! [automount:1353] CPU 0: Modules linked in: nfs nfsd lockd nfs_acl auth_rpcgss exportfs autofs4 fuse sunrpc ipv6 dm_multipath pcspkr xen_netfront dm_snapshot dm_zero dm_mirror dm_mod xen_blkfront ext3 jbd mbcache uhci_hcd ohci_hcd ehci_hcd Pid: 1353, comm: automount Tainted: G D 2.6.25-2.fc9.x86_64.xen #1 RIP: e030:[<ffffffff8030f4d0>] [<ffffffff8030f4d0>] __get_user_8+0x20/0x2c RSP: e02b:ffff880007375b50 EFLAGS: 00010206 RAX: ffffffffff516000 RBX: 0000000000000000 RCX: 000000004054e9f0 RDX: 0000000000002600 RSI: 0000000000000000 RDI: ffff88001f442000 RBP: ffff880007375ba8 R08: ffff88001f428000 R09: ffff880001760300 R10: 0000000000000006 R11: 3462646435643930 R12: ffff88001f442000 R13: 000000004054e9f0 R14: 0000000000000000 R15: ffff88001f442000 FS: 0000000000000000(0000) GS:ffffffff805bf000(0000) knlGS:0000000000000000 CS: e033 DS: 0000 ES: 0000 CR0: 000000008005003b CR2: 000000004054e9f0 CR3: 000000001dd9e000 CR4: 0000000000000660 DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000 DR3: 0000000000000000 DR6: 00000000ffff0ff0 DR7: 0000000000000000 -- Andriy Gapon

2 2

HVM DomU on F9?
by Dustin Henning 21 May '08

21 May '08

I am currently running F7 with HVMs. I am considering trying out F8 or F9 to see if I can get them to work as well. I know F9 doesn't have a Dom0, at least not yet. I also understand that DomUs are supposed to work on it. However, I don't believe I have seen any discussion as to how that is done and/or whether or not HVM DomUs are supported in that environment. I am assuming that since the decision that F9 wouldn't include Dom0 support was relatively recent that I probably don't need to search the list archives. Can anyone tell me whether or not HVM DomUs work in F9 and whether or not there is any additional performance degradation vs running them in a Dom0? I actually prefer to run HVM DomUs in a regular kernel if possible because I have always had mouse problems in Dom0 of F7, but not in the base kernel. I understand that KVM is an option, but since Xen now supports Windows with ACPI (without the degradation that was once common in Xen and is still common in KVM because of a certain ACPI register apparently regularly polled by Windows [I would love to know what register, as I could then try to find a way to prevent Windows from doing that and KVM might perform as well as Xen]), I feel that KVM would be a big step backwards. Additionally, there are apparently more paravirtual drivers available for Windows (though they are not readily available in binary form and stable, more on that in my next message), though using std-vga in KVM is nice (with standard VESA 2.0, a generic Windows driver already exists outside of the project). Finally, assuming that HVM DomUs can be run in F9, do the old services and xm commands still exist, or will I have to learn to use virsh and make xml configs? Thanks, Dustin

3 6

Re: [Fedora-xen] Paravirtual Windows Driver Binaries?
by Emre Erenoglu 21 May '08

21 May '08

Hi Dustin, Please follow the mailing list for xen-users as there's a 0.9.0 version of the drivers. The binary releases never need compilation. I guess you are missing a point here. The only possibility that you would need to compile is if you download the daily source code from the mercurial repository, but then even it's a trivial task as you just say "bld" in the root directory of the driver source code, after installing the latest WDK. Emre On Tue, May 20, 2008 at 5:54 PM, Dustin Henning <Dustin.Henning(a)prd-inc.com> wrote: > Apparently a James Harper in Australia has been working > on PV drivers for Windows on GPL Xen for some time. I don't know if this > link will work for others, but it looks like it should (though probably not > permanently): > > > http://markmail.org/search/?q=WindowsXenPV-0.8.9.zip#query:WindowsXenPV-0.8… > > I downloaded his latest work, WindowsXenPV-0.8.9.zip, but > apparently compilation is required. I know Windows drivers and programs > don't have to be different (on account of Windows) the way they do on Linux > (because of the many different distros), so I am not sure why compilation is > required unless it is somehow related to the version of Xen running on the > host system potentially offering up paravirtualization in a different way. > I am not sure on this because the directions are slightly vague and if I > read them properly, source code for the version of Xen in question is not > mentioned. Anyway, I am curious as to whether or not anyone can point me > toward a these drivers in an already compiled state. I don't know if they > would be included in any certain repository of which I am unaware or whether > it is more likely that they simply don't exist (yet). As I am running F7, I > am also curious as to whether or not Fedora's custom Xen kernels would even > support the PV capabilities on HVMs. I say on HVMs because it is my > understanding that the Windows machine decides which drivers to load (and > which devices to detect) based on a boot switch in the VM OS and I haven't > seen anything indicating a separate Xen setup for full PV. > > I ask because I would love to download binary drivers and > have a go at this, but actually building it all is a bit more complicated > and time consuming for me (especially if I'm re-inventing the wheel), as > I've never compiled anything in Windows and don't have any of the necessary > software readily available (free to download, but I don't have any seriously > high speed access). I also don't know whether it is possible that the > Fedora team (or any third party) would have any interest in compiling these > drivers or making them available even if someone else did the work since > they are for Windows and would certainly require manual installation after > download/(yum) "install". It would be possible (though far from automated) > to mount the VM partition and copy the directory tree to it, so downloading > them directly to the Dom0 doesn't seem like a problem to me, but downloading > them on the DomU would work as well if they were available some other way. > I would love to find out that the version of Xen doesn't matter for PV > drivers (beyond being above 3.0 or whatever), that the Fedora version is > standard in regards to PV, that a compiled version is available because of > that, and that one can download it from URL X as such, but I am rarely so > lucky. > > As an interesting aside, apparently the KVM project > includes a PV driver for Windows networking, but this (not necessarily > production ready) Xen GPLPV driver package includes not only networking, but > also block devices and a service to enable the xm shutdown and xm restart > commands. Anyway, I welcome all input. Thanks, > > Dustin > > -- > Fedora-xen mailing list > Fedora-xen(a)redhat.com > https://www.redhat.com/mailman/listinfo/fedora-xen > > -- Emre Erenoglu erenoglu(a)gmail.com

2 1

Paravirtual Windows Driver Binaries?
by Dustin Henning 20 May '08

20 May '08

Apparently a James Harper in Australia has been working on PV drivers for Windows on GPL Xen for some time. I don't know if this link will work for others, but it looks like it should (though probably not permanently): http://markmail.org/search/?q=WindowsXenPV-0.8.9.zip#query:WindowsXenPV-0.8. 9.zip+page:1+mid:gmf4kraasqdxxamm+state:facets I downloaded his latest work, WindowsXenPV-0.8.9.zip, but apparently compilation is required. I know Windows drivers and programs don't have to be different (on account of Windows) the way they do on Linux (because of the many different distros), so I am not sure why compilation is required unless it is somehow related to the version of Xen running on the host system potentially offering up paravirtualization in a different way. I am not sure on this because the directions are slightly vague and if I read them properly, source code for the version of Xen in question is not mentioned. Anyway, I am curious as to whether or not anyone can point me toward a these drivers in an already compiled state. I don't know if they would be included in any certain repository of which I am unaware or whether it is more likely that they simply don't exist (yet). As I am running F7, I am also curious as to whether or not Fedora's custom Xen kernels would even support the PV capabilities on HVMs. I say on HVMs because it is my understanding that the Windows machine decides which drivers to load (and which devices to detect) based on a boot switch in the VM OS and I haven't seen anything indicating a separate Xen setup for full PV. I ask because I would love to download binary drivers and have a go at this, but actually building it all is a bit more complicated and time consuming for me (especially if I'm re-inventing the wheel), as I've never compiled anything in Windows and don't have any of the necessary software readily available (free to download, but I don't have any seriously high speed access). I also don't know whether it is possible that the Fedora team (or any third party) would have any interest in compiling these drivers or making them available even if someone else did the work since they are for Windows and would certainly require manual installation after download/(yum) "install". It would be possible (though far from automated) to mount the VM partition and copy the directory tree to it, so downloading them directly to the Dom0 doesn't seem like a problem to me, but downloading them on the DomU would work as well if they were available some other way. I would love to find out that the version of Xen doesn't matter for PV drivers (beyond being above 3.0 or whatever), that the Fedora version is standard in regards to PV, that a compiled version is available because of that, and that one can download it from URL X as such, but I am rarely so lucky. As an interesting aside, apparently the KVM project includes a PV driver for Windows networking, but this (not necessarily production ready) Xen GPLPV driver package includes not only networking, but also block devices and a service to enable the xm shutdown and xm restart commands. Anyway, I welcome all input. Thanks, Dustin

1 0

Update of dom0 posible?
by Maximilian Freisinger 19 May '08

19 May '08

Hello, as there is no dom0 support in FC9, i am about to use FC8 as dom0 with FC9 as domU. My question now is: If there should be an update for dom0 support in FC9, will it be possible to install a FC9 dom0 system and move all domUs to that system, on the same machine? My idea was to setup a functional dom0 and move all configs to that system, but as FC8 is xen 3.1.2 and anything else would at least be 3.2.x it would be interesting to know if a port is possible. dc _________________________________________________________________ Give to a good cause with every e-mail. Join the i’m Initiative from Microsoft. http://im.live.com/Messenger/IM/Join/Default.aspx?souce=EML_WL_ GoodCause

2 1

Dump guest memory in Xen?
by Richard W.M. Jones 19 May '08

19 May '08

I might be missing something obvious but ... QEMU has a nice 'memsave' command which lets you save/dump parts of guest memory to a file. Is there a similar feature in Xen? (Note I don't want to dump all memory to a core file, just snoop on small sections of memory). Rich. -- Richard Jones, Emerging Technologies, Red Hat http://et.redhat.com/~rjones virt-p2v converts physical machines to virtual machines. Boot with a live CD or over the network (PXE) and turn machines into Xen guests. http://et.redhat.com/~rjones/virt-p2v

2 1

backup domU
by Mathieu Kretchner 19 May '08

19 May '08

Hello all, I'm sure it's a frequent question asked a lot of time, But is it possible to backup domU without stopping it ? Else could you please give me some url talking about the subject ? (it's difficult to find something in the list archives)

5 10

Problem adding new domu
by Greg Hellings 16 May '08

16 May '08

I'm trying to create a paravirtual domu on a system that I've already got an hvm domu on using Virt-Manager. I am choosing Shared physical device on the network page, and 'peth0 (Bridge eth0)' is the only selectable option for Device. I choose this, and then I get the following error: Unable to complete install: 'virDomainCreatelinux() failed POST operation failed: (xend.err 'Device 0 (vif) could not be connected. Hotplug scripts not working. What am I doing wrong? -- Greg

1 0

Re: [Fedora-xen] Plans for paravirt_ops kernel-xen
by Daniel Spies 16 May '08

16 May '08

Hi there! It's May now, and today is the day for Fedora 9 to be released. So any news on F10 dom0 support? Is this already implemented in the meantime? I don't want to nag anyone, don't get me wrong, please... I am using Kubuntu 8.04 ATM, and am facing loads of troubles with Xen. Now I wanted to give F9 a try, but it turns out that there's no dom0 support... So I would like to give F10/rawhide a try, and maybe get things working again... :) Thanks, Daniel Pasi Kärkkäinen wrote: > > On Sun, Mar 09, 2008 at 02:09:26PM +0200, Pasi Kärkkäinen wrote: >> On Sun, Mar 09, 2008 at 08:36:39AM -0300, Itamar - IspBrasil wrote: >> > I belive at the end of this year. >> > >> >> Hmm.. I think it's sooner.. >> >> Mark wrote like this: >> >> "Work will continue on the Dom0 paravirt_ops effort for Fedora 10 and we >> hope to introduce the first build to rawhide soon after Fedora 9 been >> branched." >> >> I was just wondering when that actually is.. is it when F9 is released >> (29 >> April 2008) or sooner, perhaps after F9 beta freeze? >> > > Actually looking at this: > http://fedoraproject.org/wiki/Releases/9/Schedule > > 8 April 2008 "Branch all packages for Fedora 9" > > So I guess that's the answer.. :) > > -- Pasi > >> >> > >> > > >> > >Yep. Thanks for the update! >> > > >> > >When is fedora 10 rawhide expected to open? Meaning when we can start >> > >testing kernel-xen with dom0 support :) >> > > >> > >-- Pasi >> > > >> > > >> > >> > >> > > -- > Fedora-xen mailing list > Fedora-xen(a)redhat.com > https://www.redhat.com/mailman/listinfo/fedora-xen > > -- View this message in context: http://www.nabble.com/Plans-for-paravirt_ops-kernel-xen-tp15876181p17205184… Sent from the Fedora Xen mailing list archive at Nabble.com.

4 5

Question on Windows HVM Domain
by Robert Locke 16 May '08

16 May '08

Bit of a n00b question.... I am currently running F8-Dom0 with an HVM-based Windows XP DomU. I tend to run it from virt-manager the occasional times I use it. Given that F9-Dom0 support is still a work-in-progress, I started reading about KVM existing in the "regular" kernel. Since the Dom0 is my main "work" platform, I am considering upgrading to F9, but, losing Xen. The question becomes, can I run the existing Windows XP instance under virt-manager via KVM? BTW, the disk for Windows XP is an LV, and I would prefer to continue NAT'ing it to the outside, if that matters.... Some of the doc's on fedoraproject.org started identifying parity between KVM and Qemu (which I thought is what is used for HVM).... Am I totally insane, or on to a potential solution that lets me move my Dom0 to F9? Thanks for any pointers/help, --Rob

3 2

[ANNOUNCE] virt-df 2.1.0 - a 'df' tool for virtual guests
by Richard W.M. Jones 15 May '08

15 May '08

I'm pleased to announce the most recent release of virt-df (2.1.0). Virt-df is 'df' for virtual guests. Run the program on the host / dom0 to display disk space used and available on all partitions on all guests. You don't need to run any sort of program/agent within the guest. Home page: http://et.redhat.com/~rjones/virt-df/ Source/binaries: http://et.redhat.com/~rjones/virt-df/files/ Developer repository: http://hg.et.redhat.com/virt/applications/virt-df--devel This version supports most common filesystems and partitioning schemes, including: - Linux ext2/3 - DOS FAT32 - Windows NTFS - Linux LVM2 (volume groups and logical volumes) - Primary and extended disk partitions - Linux swap - Linux suspend partition You can run it in your host / dom0 to display guest filesystems: # virt-df -c qemu:///system -h Filesystem Size Used Available Type rhel51x32kvm:hda1 96.8 MiB 14.6 MiB 82.2 MiB Linux ext2/3 rhel51x32kvm:VolGroup00/LogVol00 6.4 GiB 3.6 GiB 2.8 GiB Linux ext2/3 rhel51x32kvm:VolGroup00/LogVol01 992.0 MiB Linux swap You can also run it on general disk images, or disk devices: # virt-df -t /dev/sda Filesystem 1K-blocks Used Available Type /dev/sda:hda1 25599996 11309448 14290552 Windows NTFS /dev/sda:hda2 992016 93772 898244 Linux ext2/3 /dev/sda:F9VG/F9Root 23316072 7818164 15497908 Linux ext2/3 /dev/sda:F9VG/F9Swap 1015808 Linux swap /dev/sda:RHEL51VG/RHEL51Root 22382184 7796640 14585544 Linux ext2/3 /dev/sda:RHEL51VG/RHEL51Swap 2031616 Linux swap /dev/sda:VolGroup/FAT32Test 916736 4 914676 DOS/Windows You can write the output to a CSV file (use --csv option) in order to import the data easily into spreadsheets and databases. Included also is an experimental command line tool called 'diskzip' which intelligently compresses disk images by leaving out the bits which aren't actually used in the filesystems / partitions / volume groups contained within. Rich. -- Richard Jones, Emerging Technologies, Red Hat http://et.redhat.com/~rjones virt-p2v converts physical machines to virtual machines. Boot with a live CD or over the network (PXE) and turn machines into Xen guests. http://et.redhat.com/~rjones/virt-p2v

2 3

Paravirtualization with Fedora 7 in Pentium IV
by Gastón Keller 15 May '08

15 May '08

Hello, everybody. I'm planning to play with Xen (version 3.1.2) in a machine with a Pentium IV and 2 GB of memory. Fedora 7 is the installed OS (kernel 2.6.21-7.fc7xen), so I guess I'm going to use Fedora 7 for guests as well (can I use Fedora 8 for guests?). I'm considering using a second disk of 140 GB with LVM, in order to change the disk space allocated to the guests in case I need to. My next consideration is how to deal with the installation tree. I think that downloading the files from a repository each time I want to create a new guest would be kind of silly, so I'm considering the suggestion in the Fedora7VirtQuickStart [0] of downloading it once and making it available from the host OS through NFS... or maybe through HTTP using a web server (this last option is just a sudden thought without much evaluation of its value =P ). Any suggestion or comment? I mean. I might be doing something stupid due to my lack of experience. :) Thanks, Gaston PS: I'm trying to give all the info possible, not that is actually needed. [0] http://fedoraproject.org/wiki/Docs/Fedora7VirtQuickStart#head-546553e2f5a1e… -- La única verdad es la realidad.

2 1

Fedora kernel-xen pv_ops plans after F9 release (F10 rawhide)
by Pasi Kärkkäinen 13 May '08

13 May '08

Hello list! Fedora 9 will be released shortly, so I thought I'll ask about the kernel-xen pv_ops plans and status.. would be nice to get some short progress report. Are these wiki pages up to date? http://fedoraproject.org/wiki/Features/XenPvops http://fedoraproject.org/wiki/Features/XenPvopsDom0 Waiting for dom0 capable kernel-xen to test :) -- Pasi

2 2

Custom bridge script; upgrade from Fedora 6 to Fedora 8
by Ask Bjørn Hansen 07 May '08

07 May '08

Hi, Since upgrading from Fedora 6 to Fedora 8 my custom bridge script stopped working. I need to have both eth0 and eth1 bridged into each XM. My xen machines are just defined in old-style /etc/xen/foo configuration files, like vif = [ 'mac=00:16:3e:42:08:1d, bridge=xenbr0', 'mac=00:16:3e: 42:08:1e, bridge=xenbr1' ] I've changed that to vif = [ 'mac=00:16:3e:42:08:1d, bridge=eth0', 'mac=00:16:3e:42:08:1e, bridge=eth1' ] but when I do a "xm create foo" I still end up with the new interfaces on the wrong bridge! # brctl show bridge name bridge id STP enabled interfaces eth0 8000.0030488c8a26 no peth0 eth1 8000.0030488c8a27 no peth1 xenbr0 8000.feffffffffff no vif1.0 xenbr1 8000.feffffffffff no vif1.1 If I manually move the vifX.X interfaces to the eth0/eth1 bridges then everything works. This is my custom bridge script: #!/bin/sh dir=$(dirname "$0") "$dir/network-bridge" "$@" netdev=eth1 bridge=eth1 "$dir/network-bridge" "$@" netdev=eth0 bridge=eth0 Any ideas? - ask -- http://develooper.com/ - http://askask.com/

3 5

2025

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

2007

2006

2005

xen May 2008