I''ve spent some time this weekend attempting to get domU''s installed on a dom0 that''s Indiana Preview 2 (which is basically snv_79b); system is a Toshiba M5, latest BIOS, and the dom0 is booted and working just fine. On my various attempts with virt-install, it never appears to boot the guest ISO, just sits for a couple of minutes and fails. The last portion of xend.log in every case is a sequence such as below. I''t possible (probable, even) that this is problem with Indiana, but I''d like to get to the bottom of it so we can correct it (and I need some domU''s for things I''m doing). Ideas on what to look at? Dave [2008-02-10 12:52:26 xend 3669] DEBUG (XendDomain:431) Adding Domain: 23 [2008-02-10 12:52:26 xend 3669] DEBUG (DevController:149) Waiting for devices vif. [2008-02-10 12:52:26 xend.XendDomainInfo 3669] DEBUG (XendDomainInfo:824) XendDomainInfo.handleShutdownWatch [2008-02-10 12:52:26 xend 3669] DEBUG (DevController:154) Waiting for 0. [2008-02-10 12:52:26 xend 3669] DEBUG (DevController:561) hotplugStatusCallback /local/domain/0/backend/vif/23/0/hotplug-status. [2008-02-10 12:54:06 xend 3669] DEBUG (DevController:549) destroyCallback /local/domain/23/device/vif/0. [2008-02-10 12:54:16 xend 3669] ERROR (SrvBase:88) Request wait_for_devices failed. Traceback (most recent call last): File "/usr/lib/python2.4/vendor-packages/xen/web/SrvBase.py", line 85, in perform return op_method(op, req) File "/usr/lib/python2.4/vendor-packages/xen/xend/server/SrvDomain.py", line 85, in op_wait_for_devices return self.dom.waitForDevices() File "/usr/lib/python2.4/vendor-packages/xen/xend/XendDomainInfo.py", line 518, in waitForDevices self.getDeviceController(devclass).waitForDevices() File "/usr/lib/python2.4/vendor-packages/xen/xend/server/DevController.py", line 150, in waitForDevices return map(self.waitForDevice, self.deviceIDs()) File "/usr/lib/python2.4/vendor-packages/xen/xend/server/DevController.py", line 159, in waitForDevice self.destroyDevice(devid, False) File "/usr/lib/python2.4/vendor-packages/xen/xend/server/DevController.py", line 237, in destroyDevice raise EnvironmentError EnvironmentError [2008-02-10 12:54:16 xend.XendDomainInfo 3669] DEBUG (XendDomainInfo:1519) XendDomainInfo.destroy: domid=23 [2008-02-10 12:54:16 xend.XendDomainInfo 3669] DEBUG (XendDomainInfo:1527) XendDomainInfo.destroyDomain(23) [2008-02-10 12:54:17 xend 3669] DEBUG (DevController:549) destroyCallback /local/domain/23/device/vif/0. [2008-02-10 12:54:17 xend 3669] DEBUG (DevController:557) destroyCallback /local/domain/23/device/vif/0 is destroyed This message posted from opensolaris.org
I manage to install Indiana2 HVM using phyton config file, pls see attached. Pls change boot=d to boot from CD iso file. However the gdm can not start Xorg, you need to create xorg.conf file or start vncserver so that you can run gui-install from xterm. You can not use virt-install paravirtualized since the i86xpv kernel is not in miniroot file. Rgds, Andre W. Dave Miner wrote:> I''ve spent some time this weekend attempting to get domU''s installed on a dom0 that''s Indiana Preview 2 (which is basically snv_79b); system is a Toshiba M5, latest BIOS, and the dom0 is booted and working just fine. On my various attempts with virt-install, it never appears to boot the guest ISO, just sits for a couple of minutes and fails. The last portion of xend.log in every case is a sequence such as below. I''t possible (probable, even) that this is problem with Indiana, but I''d like to get to the bottom of it so we can correct it (and I need some domU''s for things I''m doing). Ideas on what to look at? > > Dave > > [2008-02-10 12:52:26 xend 3669] DEBUG (XendDomain:431) Adding Domain: 23 > [2008-02-10 12:52:26 xend 3669] DEBUG (DevController:149) Waiting for devices vif. > [2008-02-10 12:52:26 xend.XendDomainInfo 3669] DEBUG (XendDomainInfo:824) XendDomainInfo.handleShutdownWatch > [2008-02-10 12:52:26 xend 3669] DEBUG (DevController:154) Waiting for 0. > [2008-02-10 12:52:26 xend 3669] DEBUG (DevController:561) hotplugStatusCallback /local/domain/0/backend/vif/23/0/hotplug-status. > [2008-02-10 12:54:06 xend 3669] DEBUG (DevController:549) destroyCallback /local/domain/23/device/vif/0. > [2008-02-10 12:54:16 xend 3669] ERROR (SrvBase:88) Request wait_for_devices failed. > Traceback (most recent call last): > File "/usr/lib/python2.4/vendor-packages/xen/web/SrvBase.py", line 85, in perform > return op_method(op, req) > File "/usr/lib/python2.4/vendor-packages/xen/xend/server/SrvDomain.py", line 85, in op_wait_for_devices > return self.dom.waitForDevices() > File "/usr/lib/python2.4/vendor-packages/xen/xend/XendDomainInfo.py", line 518, in waitForDevices > self.getDeviceController(devclass).waitForDevices() > File "/usr/lib/python2.4/vendor-packages/xen/xend/server/DevController.py", line 150, in waitForDevices > return map(self.waitForDevice, self.deviceIDs()) > File "/usr/lib/python2.4/vendor-packages/xen/xend/server/DevController.py", line 159, in waitForDevice > self.destroyDevice(devid, False) > File "/usr/lib/python2.4/vendor-packages/xen/xend/server/DevController.py", line 237, in destroyDevice > raise EnvironmentError > EnvironmentError > [2008-02-10 12:54:16 xend.XendDomainInfo 3669] DEBUG (XendDomainInfo:1519) XendDomainInfo.destroy: domid=23 > [2008-02-10 12:54:16 xend.XendDomainInfo 3669] DEBUG (XendDomainInfo:1527) XendDomainInfo.destroyDomain(23) > [2008-02-10 12:54:17 xend 3669] DEBUG (DevController:549) destroyCallback /local/domain/23/device/vif/0. > [2008-02-10 12:54:17 xend 3669] DEBUG (DevController:557) destroyCallback /local/domain/23/device/vif/0 is destroyed > > > This message posted from opensolaris.org > _______________________________________________ > xen-discuss mailing list > xen-discuss@opensolaris.org >
Dave Miner wrote:> I''ve spent some time this weekend attempting to get domU''s installed on a > dom0 that''s Indiana Preview 2 (which is basically snv_79b); system is a > Toshiba M5, latest BIOS, and the dom0 is booted and working just fine. > On my various attempts with virt-install, it never appears to boot the > guest ISO, just sits for a couple of minutes and fails. The last portion > of xend.log in every case is a sequence such as below. I''t possible > (probable, even) that this is problem with Indiana, but I''d like to get > to the bottom of it so we can correct it (and I need some domU''s for things > I''m doing). Ideas on what to look at? > > Dave > > [2008-02-10 12:52:26 xend 3669] DEBUG (XendDomain:431) Adding Domain: 23 > [2008-02-10 12:52:26 xend 3669] DEBUG (DevController:149) Waiting for devices vif. > [2008-02-10 12:52:26 xend.XendDomainInfo 3669] DEBUG (XendDomainInfo:824) XendDomainInfo.handleShutdownWatch > [2008-02-10 12:52:26 xend 3669] DEBUG (DevController:154) Waiting for 0. > [2008-02-10 12:52:26 xend 3669] DEBUG (DevController:561) hotplugStatusCallback /local/domain/0/backend/vif/23/0/hotplug-status. > [2008-02-10 12:54:06 xend 3669] DEBUG (DevController:549) destroyCallback /local/domain/23/device/vif/0. > [2008-02-10 12:54:16 xend 3669] ERROR (SrvBase:88) Request wait_for_devices failed. > Traceback (most recent call last): > File "/usr/lib/python2.4/vendor-packages/xen/web/SrvBase.py", line 85, in perform > return op_method(op, req) > File "/usr/lib/python2.4/vendor-packages/xen/xend/server/SrvDomain.py", line 85, in op_wait_for_devices > return self.dom.waitForDevices() > File "/usr/lib/python2.4/vendor-packages/xen/xend/XendDomainInfo.py", line 518, in waitForDevices > self.getDeviceController(devclass).waitForDevices() > File "/usr/lib/python2.4/vendor-packages/xen/xend/server/DevController.py", line 150, in waitForDevices > return map(self.waitForDevice, self.deviceIDs()) > File "/usr/lib/python2.4/vendor-packages/xen/xend/server/DevController.py", line 159, in waitForDevice > self.destroyDevice(devid, False) > File "/usr/lib/python2.4/vendor-packages/xen/xend/server/DevController.py", line 237, in destroyDevice > raise EnvironmentError > EnvironmentError > [2008-02-10 12:54:16 xend.XendDomainInfo 3669] DEBUG (XendDomainInfo:1519) XendDomainInfo.destroy: domid=23 > [2008-02-10 12:54:16 xend.XendDomainInfo 3669] DEBUG (XendDomainInfo:1527) XendDomainInfo.destroyDomain(23) > [2008-02-10 12:54:17 xend 3669] DEBUG (DevController:549) destroyCallback /local/domain/23/device/vif/0. > [2008-02-10 12:54:17 xend 3669] DEBUG (DevController:557) destroyCallback /local/domain/23/device/vif/0 is destroyedIt seems to have problems configuring the domU''s nic device. Did you check /var/log/xen/xpvd-event.log ? The shell script that sets up a domU nic (/usr/lib/xen/scripts/vif-vnic) should write some trace information to that /var/log/xen/xpvd-event.log file. Things that can go wrong with the nic: - no GLDv3 nic found ("legacy" nic devices showing up in "dladm show-link" output cannot be used with xen domU) - multiple non-legacy nics found. /usr/lib/xen/scripts/vif-vnic picks one as the default, but it might not be the one that is connected to your network. - WLAN GLDv3 nics cannot be used as a backend nic device for xVM/domU xenstore-read can be used to read the default nic used with domU: /usr/lib/xen/bin/xenstore-read device-misc/vif/default-nic When this isn''t set, it looks at the config/default-nic property for smf service svc:/system/xvm/xend:default svcprop -p config/default-nic svc:/system/xvm/xend:default And when that property isn''t set, it uses the first non-legacy nic from "dladm show-link" output.
Andre Wenas wrote:> I manage to install Indiana2 HVM using phyton config file, pls see > attached. Pls change boot=d to boot from CD iso file. > However the gdm can not start Xorg, you need to create xorg.conf file or > start vncserver so that you can run gui-install from xterm. > > You can not use virt-install paravirtualized since the i86xpv kernel is > not in miniroot file.Right. The PV kernel is in the root of the iso, but it did not get put in the ramdisk. Also, the ramdisk name is changed from a standard Nevada iso so virt-install wouldn''t find it anyway. This would have been the py file to use to install it manually. Also, we don''t have zfs boot support yet for PV guests. We are working on this now. -- name = "indiana-pv-install" vcpus = 1 memory = "1024" bootloader = "/usr/lib/xen/bin/pygrub" kernel = "/platform/i86xpv/kernel/amd64/unix" ramdisk = "/boot/x86.microroot" extra = "/platform/i86xpv/kernel/amd64/unix -k -B console=ttya,livemode-text" disk = [''file:/tank/guests/install/indiana/in-test-199.iso,6:cdrom,r'', ''file:/tank/guests/install/indiana/disk.img,0,w''] vif = [''''] on_shutdown = "destroy" on_reboot = "restart" on_crash = "preserve" To get X working in HVM, see the following instructions. -- Start the domain using the install.py at the end of this e-mail (edit it for your paths) # xm create install.py In another terminal, telnet to the domains "serial" port # telnet localhost 4444 In the HVM console in the grub menu, select text, edit the kernel line and add ",console=ttya" after livemode. Boot. In the serial console, login user: jack passwd: jack # su - (password is opensolaris) # /usr/X11/bin/Xorg -configure edit /root/xorg.conf.new and change the adapter driver from "cirrius" to "vesa". Start X in the console from the shell in the serial port. # /usr/X11/bin/xinit /usr/bin/dbus-launch /usr/bin/gnome-session -- /usr/X11/bin/Xorg -config /root/xorg.conf.new :0 bring up a gnome terminal and launch the installer. # install-lan --- : alpha[1]#; cat install.py name = "indiana-install" vcpus=1 memory = 1024 shadow_memory = 8 vif = [ ''type=ioemu'' ] disk = [''file:/tank/guests/install/indiana/disk.img,hda,w'', ''file:/tank/guests/install/indiana/in-test-199.iso,hdd:cdrom,r''] builder=''hvm'' kernel = "/usr/lib/xen/boot/hvmloader" on_poweroff = ''destroy'' on_reboot = ''restart'' on_crash = ''preserve'' boot=''d'' acpi=0 apic=0 sdl=0 vnc=1 vnclisten="0.0.0.0" vncpasswd=''password'' vncconsole=1 nographic=0 stdvga=0 serial=''telnet:0.0.0.0:4444,server,nowait'' usb=1 usbdevice="tablet" soundhw=''es1370'' import os, re arch = os.uname()[4] if re.search(''64'', arch): arch_libdir = ''lib64'' else: arch_libdir = ''lib'' device_model = ''/usr/'' + arch_libdir + ''/xen/bin/qemu-dm''
In my case, after installation, I need to add atapi-cd-dma-enabled=0 and atapi-other-dma-enabled=0 in grub. To boot 64bit kernel, you need to speficy amd64 otherwise it will boot 32bit. Here is my grub config: -bash-3.2# tail /rpool/boot/grub/menu.lst default 0 #---------- ADDED BY BOOTADM - DO NOT EDIT ---------- title OpenSolaris Developer Preview 2 snv_79b X86 64bit kernel$ /platform/i86pc/kernel/amd64/unix -B $ZFS-BOOTFS,atapi-cd-dma-enabled=0,atapi,other-dma-enabled=0 module$ /platform/i86pc/amd64/boot_archive #---------------------END BOOTADM-------------------- title OpenSolaris Developer Preview 2 snv_79b X86 32bit kernel$ /platform/i86pc/kernel/$ISADIR/unix -B $ZFS-BOOTFS,atapi-cd-dma-enabled=0,atapi-other-dma-enabled=0 module$ /platform/i86pc/$ISADIR/boot_archive #---------------------END BOOTADM-------------------- Rgds, Andre W. Mark Johnson wrote:> > > Andre Wenas wrote: >> I manage to install Indiana2 HVM using phyton config file, pls see >> attached. Pls change boot=d to boot from CD iso file. >> However the gdm can not start Xorg, you need to create xorg.conf file >> or start vncserver so that you can run gui-install from xterm. >> >> You can not use virt-install paravirtualized since the i86xpv kernel >> is not in miniroot file. > > Right. The PV kernel is in the root of the iso, but > it did not get put in the ramdisk. Also, the ramdisk > name is changed from a standard Nevada iso so virt-install > wouldn''t find it anyway. This would have been the py > file to use to install it manually. > > Also, we don''t have zfs boot support yet for PV > guests. We are working on this now. > > > -- > name = "indiana-pv-install" > vcpus = 1 > memory = "1024" > > bootloader = "/usr/lib/xen/bin/pygrub" > kernel = "/platform/i86xpv/kernel/amd64/unix" > ramdisk = "/boot/x86.microroot" > extra = "/platform/i86xpv/kernel/amd64/unix -k -B > console=ttya,livemode-text" > > disk = [''file:/tank/guests/install/indiana/in-test-199.iso,6:cdrom,r'', > ''file:/tank/guests/install/indiana/disk.img,0,w''] > vif = [''''] > > on_shutdown = "destroy" > on_reboot = "restart" > on_crash = "preserve" > > > > To get X working in HVM, see the following instructions. > > -- > > > Start the domain using the install.py at the > end of this e-mail (edit it for your paths) > > # xm create install.py > > > > In another terminal, telnet to the domains "serial" > port > # telnet localhost 4444 > > In the HVM console in the grub menu, select text, > edit the kernel line and add ",console=ttya" after > livemode. Boot. > > > In the serial console, login > user: jack > passwd: jack > > # su - > (password is opensolaris) > > # /usr/X11/bin/Xorg -configure > > edit /root/xorg.conf.new and change the > adapter driver from "cirrius" to "vesa". > > Start X in the console from the shell in > the serial port. > > # /usr/X11/bin/xinit /usr/bin/dbus-launch /usr/bin/gnome-session -- > /usr/X11/bin/Xorg -config /root/xorg.conf.new :0 > > bring up a gnome terminal and launch the installer. > > # install-lan > > > > --- > > : alpha[1]#; cat install.py > > name = "indiana-install" > vcpus=1 > memory = 1024 > shadow_memory = 8 > > vif = [ ''type=ioemu'' ] > > disk = [''file:/tank/guests/install/indiana/disk.img,hda,w'', > ''file:/tank/guests/install/indiana/in-test-199.iso,hdd:cdrom,r''] > > builder=''hvm'' > kernel = "/usr/lib/xen/boot/hvmloader" > > on_poweroff = ''destroy'' > on_reboot = ''restart'' > on_crash = ''preserve'' > > boot=''d'' > > acpi=0 > apic=0 > sdl=0 > vnc=1 > vnclisten="0.0.0.0" > vncpasswd=''password'' > vncconsole=1 > nographic=0 > stdvga=0 > serial=''telnet:0.0.0.0:4444,server,nowait'' > usb=1 > usbdevice="tablet" > soundhw=''es1370'' > > import os, re > arch = os.uname()[4] > if re.search(''64'', arch): > arch_libdir = ''lib64'' > else: > arch_libdir = ''lib'' > device_model = ''/usr/'' + arch_libdir + ''/xen/bin/qemu-dm''
Andre Wenas wrote:> I manage to install Indiana2 HVM using phyton config file, pls see > attached. Pls change boot=d to boot from CD iso file. > However the gdm can not start Xorg, you need to create xorg.conf file or > start vncserver so that you can run gui-install from xterm. > > You can not use virt-install paravirtualized since the i86xpv kernel is > not in miniroot file. >Indiana is the dom0, not the domU; the domU''s I tried were snv_82 or fedora8. Dave> Rgds, > Andre W. > > Dave Miner wrote: >> I''ve spent some time this weekend attempting to get domU''s installed on a dom0 that''s Indiana Preview 2 (which is basically snv_79b); system is a Toshiba M5, latest BIOS, and the dom0 is booted and working just fine. On my various attempts with virt-install, it never appears to boot the guest ISO, just sits for a couple of minutes and fails. The last portion of xend.log in every case is a sequence such as below. I''t possible (probable, even) that this is problem with Indiana, but I''d like to get to the bottom of it so we can correct it (and I need some domU''s for things I''m doing). Ideas on what to look at? >> >> Dave >> >> [2008-02-10 12:52:26 xend 3669] DEBUG (XendDomain:431) Adding Domain: 23 >> [2008-02-10 12:52:26 xend 3669] DEBUG (DevController:149) Waiting for devices vif. >> [2008-02-10 12:52:26 xend.XendDomainInfo 3669] DEBUG (XendDomainInfo:824) XendDomainInfo.handleShutdownWatch >> [2008-02-10 12:52:26 xend 3669] DEBUG (DevController:154) Waiting for 0. >> [2008-02-10 12:52:26 xend 3669] DEBUG (DevController:561) hotplugStatusCallback /local/domain/0/backend/vif/23/0/hotplug-status. >> [2008-02-10 12:54:06 xend 3669] DEBUG (DevController:549) destroyCallback /local/domain/23/device/vif/0. >> [2008-02-10 12:54:16 xend 3669] ERROR (SrvBase:88) Request wait_for_devices failed. >> Traceback (most recent call last): >> File "/usr/lib/python2.4/vendor-packages/xen/web/SrvBase.py", line 85, in perform >> return op_method(op, req) >> File "/usr/lib/python2.4/vendor-packages/xen/xend/server/SrvDomain.py", line 85, in op_wait_for_devices >> return self.dom.waitForDevices() >> File "/usr/lib/python2.4/vendor-packages/xen/xend/XendDomainInfo.py", line 518, in waitForDevices >> self.getDeviceController(devclass).waitForDevices() >> File "/usr/lib/python2.4/vendor-packages/xen/xend/server/DevController.py", line 150, in waitForDevices >> return map(self.waitForDevice, self.deviceIDs()) >> File "/usr/lib/python2.4/vendor-packages/xen/xend/server/DevController.py", line 159, in waitForDevice >> self.destroyDevice(devid, False) >> File "/usr/lib/python2.4/vendor-packages/xen/xend/server/DevController.py", line 237, in destroyDevice >> raise EnvironmentError >> EnvironmentError >> [2008-02-10 12:54:16 xend.XendDomainInfo 3669] DEBUG (XendDomainInfo:1519) XendDomainInfo.destroy: domid=23 >> [2008-02-10 12:54:16 xend.XendDomainInfo 3669] DEBUG (XendDomainInfo:1527) XendDomainInfo.destroyDomain(23) >> [2008-02-10 12:54:17 xend 3669] DEBUG (DevController:549) destroyCallback /local/domain/23/device/vif/0. >> [2008-02-10 12:54:17 xend 3669] DEBUG (DevController:557) destroyCallback /local/domain/23/device/vif/0 is destroyed >> >> >> This message posted from opensolaris.org >> _______________________________________________ >> xen-discuss mailing list >> xen-discuss@opensolaris.org >> >
Juergen Keil wrote:> Dave Miner wrote: > >> I''ve spent some time this weekend attempting to get domU''s installed on a >> dom0 that''s Indiana Preview 2 (which is basically snv_79b); system is a >> Toshiba M5, latest BIOS, and the dom0 is booted and working just fine. >> On my various attempts with virt-install, it never appears to boot the >> guest ISO, just sits for a couple of minutes and fails. The last portion >> of xend.log in every case is a sequence such as below. I''t possible >> (probable, even) that this is problem with Indiana, but I''d like to get >> to the bottom of it so we can correct it (and I need some domU''s for things >> I''m doing). Ideas on what to look at? >> >> Dave >> >> [2008-02-10 12:52:26 xend 3669] DEBUG (XendDomain:431) Adding Domain: 23 >> [2008-02-10 12:52:26 xend 3669] DEBUG (DevController:149) Waiting for devices vif. >> [2008-02-10 12:52:26 xend.XendDomainInfo 3669] DEBUG (XendDomainInfo:824) XendDomainInfo.handleShutdownWatch >> [2008-02-10 12:52:26 xend 3669] DEBUG (DevController:154) Waiting for 0. >> [2008-02-10 12:52:26 xend 3669] DEBUG (DevController:561) hotplugStatusCallback /local/domain/0/backend/vif/23/0/hotplug-status. >> [2008-02-10 12:54:06 xend 3669] DEBUG (DevController:549) destroyCallback /local/domain/23/device/vif/0. >> [2008-02-10 12:54:16 xend 3669] ERROR (SrvBase:88) Request wait_for_devices failed. >> Traceback (most recent call last): >> File "/usr/lib/python2.4/vendor-packages/xen/web/SrvBase.py", line 85, in perform >> return op_method(op, req) >> File "/usr/lib/python2.4/vendor-packages/xen/xend/server/SrvDomain.py", line 85, in op_wait_for_devices >> return self.dom.waitForDevices() >> File "/usr/lib/python2.4/vendor-packages/xen/xend/XendDomainInfo.py", line 518, in waitForDevices >> self.getDeviceController(devclass).waitForDevices() >> File "/usr/lib/python2.4/vendor-packages/xen/xend/server/DevController.py", line 150, in waitForDevices >> return map(self.waitForDevice, self.deviceIDs()) >> File "/usr/lib/python2.4/vendor-packages/xen/xend/server/DevController.py", line 159, in waitForDevice >> self.destroyDevice(devid, False) >> File "/usr/lib/python2.4/vendor-packages/xen/xend/server/DevController.py", line 237, in destroyDevice >> raise EnvironmentError >> EnvironmentError >> [2008-02-10 12:54:16 xend.XendDomainInfo 3669] DEBUG (XendDomainInfo:1519) XendDomainInfo.destroy: domid=23 >> [2008-02-10 12:54:16 xend.XendDomainInfo 3669] DEBUG (XendDomainInfo:1527) XendDomainInfo.destroyDomain(23) >> [2008-02-10 12:54:17 xend 3669] DEBUG (DevController:549) destroyCallback /local/domain/23/device/vif/0. >> [2008-02-10 12:54:17 xend 3669] DEBUG (DevController:557) destroyCallback /local/domain/23/device/vif/0 is destroyed > > > It seems to have problems configuring the domU''s nic device. > > Did you check /var/log/xen/xpvd-event.log ? > > The shell script that sets up a domU nic > (/usr/lib/xen/scripts/vif-vnic) should write some > trace information to that /var/log/xen/xpvd-event.log file. >There''s no such file, but at least that''s a clue to where to look next.> Things that can go wrong with the nic: > > - no GLDv3 nic found > ("legacy" nic devices showing up in "dladm show-link" > output cannot be used with xen domU) > > - multiple non-legacy nics found. /usr/lib/xen/scripts/vif-vnic > picks one as the default, but it might not be the one that > is connected to your network. > > - WLAN GLDv3 nics cannot be used as a backend nic device > for xVM/domU > > > > xenstore-read can be used to read the default nic used with > domU: > > /usr/lib/xen/bin/xenstore-read device-misc/vif/default-nic > > > When this isn''t set, it looks at the config/default-nic property > for smf service svc:/system/xvm/xend:default > > svcprop -p config/default-nic svc:/system/xvm/xend:default > > > And when that property isn''t set, it uses the first non-legacy > nic from "dladm show-link" output. >The nic''s are e1000g0 and wpi0. The xenstore property isn''t set, but I had previously set the xend/config/default-nic property to e1000g0 on the supposition this might be the problem. Dave
Dave Miner wrote:> Andre Wenas wrote: >> I manage to install Indiana2 HVM using phyton config file, pls see >> attached. Pls change boot=d to boot from CD iso file. >> However the gdm can not start Xorg, you need to create xorg.conf file or >> start vncserver so that you can run gui-install from xterm. >> >> You can not use virt-install paravirtualized since the i86xpv kernel is >> not in miniroot file. >> > > Indiana is the dom0, not the domU; the domU''s I tried were snv_82 or > fedora8.The following packages are required for dom0 support. Do you have all of them installed? SUNWvirtinst SUNWurlgrabber SUNWlibvirt SUNWxvmhvm SUNWxvmdom SUNWxvm SUNWgccruntime SUNWgnutls SUNWlibsdl FSWxwpft FSWxwrtl what files are in /var/log/xen? Not having a /var/log/xen/xpvd-* file is interesting. MRJ
Mark Johnson wrote:> > Dave Miner wrote: >> Andre Wenas wrote: >>> I manage to install Indiana2 HVM using phyton config file, pls see >>> attached. Pls change boot=d to boot from CD iso file. >>> However the gdm can not start Xorg, you need to create xorg.conf file or >>> start vncserver so that you can run gui-install from xterm. >>> >>> You can not use virt-install paravirtualized since the i86xpv kernel is >>> not in miniroot file. >>> >> Indiana is the dom0, not the domU; the domU''s I tried were snv_82 or >> fedora8. > > > The following packages are required for dom0 support. > Do you have all of them installed? > > SUNWvirtinst > SUNWurlgrabber > SUNWlibvirt > SUNWxvmhvm > SUNWxvmdom > SUNWxvm > SUNWgccruntime > SUNWgnutls > SUNWlibsdl > FSWxwpft > FSWxwrtl >Got all of those.> what files are in /var/log/xen? Not having a > /var/log/xen/xpvd-* file is interesting. >xend.log xend-debug.log Some qemu-dm*log files. The qemu logs all basically say domid: 22 qemu: the number of cpus is 1 net_tap_get_nic: timeout waiting for hotplug at /local/domain/0/backend/vif/22/0/hotplug-status net_tap_init: nic = NULL, setphysaddr = 0 Could not initialize device ''tap'' which seems consistent with the xend.log entries I sent initially. Dave
On Mon, Feb 11, 2008 at 12:09:22PM -0500, Dave Miner wrote:> domid: 22 > qemu: the number of cpus is 1 > net_tap_get_nic: timeout waiting for hotplug at > /local/domain/0/backend/vif/22/0/hotplug-status > net_tap_init: nic = NULL, setphysaddr = 0 > Could not initialize device ''tap''What does: syseventadm list say? john
John Levon wrote:> On Mon, Feb 11, 2008 at 12:09:22PM -0500, Dave Miner wrote: > >> domid: 22 >> qemu: the number of cpus is 1 >> net_tap_get_nic: timeout waiting for hotplug at >> /local/domain/0/backend/vif/22/0/hotplug-status >> net_tap_init: nic = NULL, setphysaddr = 0 >> Could not initialize device ''tap'' > > What does: > > syseventadm list > > say? >That looks like an excellent question. It says: vendor=SUNW publisher=pcie_pci class=EC_dr subclass=ESC_dr_req /usr/lib/pci/pcidr class=$class subclass=$subclass publisher=$publisher dr_request_type=$dr_request_type dr_ap_id=$dr_ap_id vendor=SUNW publisher=px_pci class=EC_dr subclass=ESC_dr_req /usr/lib/pci/pcidr class=$class subclass=$subclass publisher=$publisher dr_request_type=$dr_request_type dr_ap_id=$dr_ap_id vendor=SUNW publisher=pxb_bcm class=EC_dr subclass=ESC_dr_req /usr/lib/pci/pcidr class=$class subclass=$subclass publisher=$publisher dr_request_type=$dr_request_type dr_ap_id=$dr_ap_id vendor=SUNW publisher=pxb_plx class=EC_dr subclass=ESC_dr_req /usr/lib/pci/pcidr class=$class subclass=$subclass publisher=$publisher dr_request_type=$dr_request_type dr_ap_id=$dr_ap_id Looks like we''re missing these, based on looking at snv_81 systems: class=EC_xendev /usr/lib/xen/scripts/xpvd-event action=$subclass domain=$domain vdev=$vdev device=$device devclass=$devclass febe=$fob class=EC_xpvsys /usr/lib/xen/scripts/xpvsys-event subclass=$subclass shutdown=$shutdown Right? Dave
David.Comay@Sun.COM
2008-Feb-11 18:39 UTC
Re: Unable to install domU''s on Indiana preview 2
> Looks like we''re missing these, based on looking at snv_81 systems: > > class=EC_xendev /usr/lib/xen/scripts/xpvd-event action=$subclass > domain=$domain vdev=$vdev device=$device devclass=$devclass febe=$fob > class=EC_xpvsys /usr/lib/xen/scripts/xpvsys-event subclass=$subclass > shutdown=$shutdownYes, unfortunately this is an area where I suspect a new IPS action will be required as SUNWxvmdomu''s postinstall script does the sysevent registration. dsc
David.Comay@Sun.COM wrote:>> Looks like we''re missing these, based on looking at snv_81 systems: >> >> class=EC_xendev /usr/lib/xen/scripts/xpvd-event action=$subclass >> domain=$domain vdev=$vdev device=$device devclass=$devclass febe=$fob >> class=EC_xpvsys /usr/lib/xen/scripts/xpvsys-event subclass=$subclass >> shutdown=$shutdown > > Yes, unfortunately this is an area where I suspect a new IPS action > will be required as SUNWxvmdomu''s postinstall script does the sysevent > registration. >And once I added those, the install of an snv_82 domU starts up. Thanks much, John! Dave
David.Comay@Sun.COM wrote:>> Looks like we''re missing these, based on looking at snv_81 systems: >> >> class=EC_xendev /usr/lib/xen/scripts/xpvd-event action=$subclass >> domain=$domain vdev=$vdev device=$device devclass=$devclass febe=$fob >> class=EC_xpvsys /usr/lib/xen/scripts/xpvsys-event subclass=$subclass >> shutdown=$shutdown > > Yes, unfortunately this is an area where I suspect a new IPS action > will be required as SUNWxvmdomu''s postinstall script does the sysevent > registration. > > dscYeah, I filed 509 to cover this. I''ll add a release note as well. Dave
Hi,>>And once I added those, the install of an snv_82 domU starts up.How? (I''m lazy today ....) regards Bernd Dave Miner wrote:> David.Comay@Sun.COM wrote: > >>> Looks like we''re missing these, based on looking at snv_81 systems: >>> >>> class=EC_xendev /usr/lib/xen/scripts/xpvd-event action=$subclass >>> domain=$domain vdev=$vdev device=$device devclass=$devclass febe=$fob >>> class=EC_xpvsys /usr/lib/xen/scripts/xpvsys-event subclass=$subclass >>> shutdown=$shutdown >>> >> Yes, unfortunately this is an area where I suspect a new IPS action >> will be required as SUNWxvmdomu''s postinstall script does the sysevent >> registration. >> >> > > And once I added those, the install of an snv_82 domU starts up. Thanks > much, John! > > Dave > > _______________________________________________ > xen-discuss mailing list > xen-discuss@opensolaris.org > >-- Bernd Schemmer, Frankfurt am Main, Germany http://home.arcor.de/bnsmb/index.html M s temprano que tarde el mundo cambiar . Fidel Castro
Bernd Schemmer wrote:> Hi, > >>> And once I added those, the install of an snv_82 domU starts up. > > > How? (I''m lazy today ....) ># pfexec syseventadm add -c EC_xendev \ /usr/lib/xen/scripts/xpvd-event ''action=$subclass'' \ ''domain=$domain'' ''vdev=$vdev'' ''device=$device'' \ ''devclass=$devclass'' ''febe=$fob'' # pfexec syseventadm add -c EC_xpvsys \ /usr/lib/xen/scripts/xpvsys-event ''subclass=$subclass'' \ ''shutdown=$shutdown'' Dave