Daniel Bareiro
2008-Jun-01 15:46 UTC
[xen-discuss] OpenSolaris 2008.05 domU on Debian dom0
Hi all! I am trying to install OpenSolaris 2008.05 PV domU on Debian GNU/Linux dom0 follow this [1] steps. Xen version is 3.0.3-1 from Debian packages. But I got this messages during boot: Configuring devices. Mounting local partitions/cdroms May 31 11:04:29 svc.startd[7]: svc:/system/coreadm:default: Method "/lib/svc/method/svc-coreadm start" failed due to signal KILL. Reading ZFS config: done. May 31 11:05:36 svc.startd[7]: svc:/system/coreadm:default: Method or service exit timed out. Killing contract 31. May 31 11:07:12 svc.startd[7]: svc:/network/routing/route:default: Method or service exit timed out. Killing contract 33. May 31 11:07:16 svc.startd[7]: svc:/network/rpc/bind:default: Couldn''t fork to execute method /lib/svc/method/rpc-bind start: Resource temporarily unavailable May 31 11:07:39 svc.startd[7]: svc:/network/routing/route:default: Method or service exit timed out. Killing contract 33. May 31 11:07:42 svc.startd[7]: svc:/system/coreadm:default: Method "/lib/svc/method/svc-coreadm refresh" failed with exit status 96. May 31 11:07:44 svc.startd[7]: svc:/network/routing/route:default: Method "/lib/svc/method/svc-route" failed due to signal KILL. May 31 11:07:44 svc.startd[7]: svc:/system/dbus:default: Method or service exit timed out. Killing contract 34. May 31 11:07:46 svc.startd[7]: svc:/system/dbus:default: Method "/lib/svc/method/svc-dbus start" failed with exit status 95. May 31 11:07:52 svc.startd[7]: system/coreadm:default misconfigured: transitioned to maintenance (see ''svcs -xv'' for details) May 31 11:08:00 svc.startd[7]: system/dbus:default failed fatally: transitioned to maintenance (see ''svcs -xv'' for details) May 31 11:08:15 svc.startd[7]: svc:/network/inetd-upgrade:default: Method or service exit timed out. Killing contract 35. May 31 11:08:25 svc.startd[7]: svc:/network/inetd-upgrade:default: Method or service exit timed out. Killing contract 35. May 31 11:08:28 svc.startd[7]: svc:/network/inetd-upgrade:default: Method "/lib/svc/method/inetd-upgrade start" failed due to signal KILL. May 31 11:12:05 svc.startd[7]: svc:/application/opengl/ogl-select:default: Method or service exit timed out. Killing contract 47. May 31 11:13:03 svc.startd[7]: svc:/application/opengl/ogl-select:default: Method or service exit timed out. Killing contract 47. May 31 11:15:53 svc.startd[7]: svc:/application/opengl/ogl-select:default: Method "/lib/svc/method/ogl-select start" failed due to signal KILL. May 31 11:15:55 svc.startd[7]: svc:/network/inetd-upgrade:default: Method "/lib/svc/method/inetd-upgrade start" failed due to signal KILL. May 31 11:15:55 svc.startd[7]: svc:/network/routing/route:default: Method or service exit timed out. Killing contract 51. May 31 11:16:01 svc.startd[7]: svc:/network/routing/route:default: Method "/lib/svc/method/svc-route" failed due to signal KILL. And I do not reach to obtain login. Some idea than can be happening? Thanks in advance. Regards, Daniel [1] http://www.mobilnews.cz/blog/?p=42 -- Daniel Bareiro - System Administrator Fingerprint: BFB3 08D6 B4D1 31B2 72B9 29CE 6696 BF1B 14E6 1D37 Powered by Debian GNU/Linux Lenny - Linux user #188.598 -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 189 bytes Desc: Digital signature URL: <http://mail.opensolaris.org/pipermail/xen-discuss/attachments/20080601/6c4546aa/attachment.bin>
I would suggest install Xen 3.1 from source (or binary) on clean Debian Instance. It works for sure. Then proceed with OpenSolaris 2008/05 following :- http://www.mobilnews.cz/blog/?p=42 --- On Sun, 6/1/08, Daniel Bareiro <daniel-listas@gmx.net> wrote: From: Daniel Bareiro <daniel-listas@gmx.net> Subject: [xen-discuss] OpenSolaris 2008.05 domU on Debian dom0 To: xen-discuss@opensolaris.org Date: Sunday, June 1, 2008, 11:46 AM Hi all! I am trying to install OpenSolaris 2008.05 PV domU on Debian GNU/Linux dom0 follow this [1] steps. Xen version is 3.0.3-1 from Debian packages. But I got this messages during boot: Configuring devices. Mounting local partitions/cdroms May 31 11:04:29 svc.startd[7]: svc:/system/coreadm:default: Method "/lib/svc/method/svc-coreadm start" failed due to signal KILL. Reading ZFS config: done. May 31 11:05:36 svc.startd[7]: svc:/system/coreadm:default: Method or service exit timed out. Killing contract 31. May 31 11:07:12 svc.startd[7]: svc:/network/routing/route:default: Method or service exit timed out. Killing contract 33. May 31 11:07:16 svc.startd[7]: svc:/network/rpc/bind:default: Couldn''t fork to execute method /lib/svc/method/rpc-bind start: Resource temporarily unavailable May 31 11:07:39 svc.startd[7]: svc:/network/routing/route:default: Method or service exit timed out. Killing contract 33. May 31 11:07:42 svc.startd[7]: svc:/system/coreadm:default: Method "/lib/svc/method/svc-coreadm refresh" failed with exit status 96. May 31 11:07:44 svc.startd[7]: svc:/network/routing/route:default: Method "/lib/svc/method/svc-route" failed due to signal KILL. May 31 11:07:44 svc.startd[7]: svc:/system/dbus:default: Method or service exit timed out. Killing contract 34. May 31 11:07:46 svc.startd[7]: svc:/system/dbus:default: Method "/lib/svc/method/svc-dbus start" failed with exit status 95. May 31 11:07:52 svc.startd[7]: system/coreadm:default misconfigured: transitioned to maintenance (see ''svcs -xv'' for details) May 31 11:08:00 svc.startd[7]: system/dbus:default failed fatally: transitioned to maintenance (see ''svcs -xv'' for details) May 31 11:08:15 svc.startd[7]: svc:/network/inetd-upgrade:default: Method or service exit timed out. Killing contract 35. May 31 11:08:25 svc.startd[7]: svc:/network/inetd-upgrade:default: Method or service exit timed out. Killing contract 35. May 31 11:08:28 svc.startd[7]: svc:/network/inetd-upgrade:default: Method "/lib/svc/method/inetd-upgrade start" failed due to signal KILL. May 31 11:12:05 svc.startd[7]: svc:/application/opengl/ogl-select:default: Method or service exit timed out. Killing contract 47. May 31 11:13:03 svc.startd[7]: svc:/application/opengl/ogl-select:default: Method or service exit timed out. Killing contract 47. May 31 11:15:53 svc.startd[7]: svc:/application/opengl/ogl-select:default: Method "/lib/svc/method/ogl-select start" failed due to signal KILL. May 31 11:15:55 svc.startd[7]: svc:/network/inetd-upgrade:default: Method "/lib/svc/method/inetd-upgrade start" failed due to signal KILL. May 31 11:15:55 svc.startd[7]: svc:/network/routing/route:default: Method or service exit timed out. Killing contract 51. May 31 11:16:01 svc.startd[7]: svc:/network/routing/route:default: Method "/lib/svc/method/svc-route" failed due to signal KILL. And I do not reach to obtain login. Some idea than can be happening? Thanks in advance. Regards, Daniel [1] http://www.mobilnews.cz/blog/?p=42 -- Daniel Bareiro - System Administrator Fingerprint: BFB3 08D6 B4D1 31B2 72B9 29CE 6696 BF1B 14E6 1D37 Powered by Debian GNU/Linux Lenny - Linux user #188.598_______________________________________________ xen-discuss mailing list xen-discuss@opensolaris.org
Hi Boris.> I would suggest install Xen 3.1 from source (or binary) on clean > Debian Instance. It works for sure. Then proceed with OpenSolaris > 2008/05 following :- http://www.mobilnews.cz/blog/?p=42At the moment I don''t have some machine for making a new installation. The rare thing is that also I read in sites of OpenSolaris that OpenSolaris 2008.05 works with Xen 3.0.4 and to earlier install [1]. Thanks for your reply, Boris. Regards, Daniel [1] http://opensolaris.org/os/community/xen/docs/linux-dom0/ -- Daniel Bareiro - System Administrator Fingerprint: BFB3 08D6 B4D1 31B2 72B9 29CE 6696 BF1B 14E6 1D37 Powered by Debian GNU/Linux Lenny - Linux user #188.598
Daniel, You are right. However, install procedure was done by myself at Xen 3.2.1 Ubuntu 8.04, CentOS 5.1 Dom0 ( thanks to Jurgen Keil and Mark Jonson) :- http://lxer.com/module/newswire/view/103407/index.html http://lxer.com/module/newswire/view/103407/index.html and at Xen 3.1 F8 Dom0 with no problems (no kernel patch needed) Thanks. -b This message posted from opensolaris.org
El domingo 08 de junio del 2008 a las 08:23:28, Boris Derzhavets escribió:> Daniel,Hi Boris.> You are right. However, install procedure was done by myself at Xen > 3.2.1 Ubuntu 8.04, CentOS 5.1 Dom0 ( thanks to Jurgen Keil and Mark > Jonson) :- http://lxer.com/module/newswire/view/103407/index.html > http://lxer.com/module/newswire/view/103407/index.html and at Xen 3.1 > F8 Dom0 with no problems (no kernel patch needed)I try installing xen-utils-3.2-1 and xen-hypervisor-3.2-1-i386 packages from backports repositories. These have Xen 3.2.0-3 version, according to I can see. This is domU''s config file: sirius:/etc/xen# cat febo.cfg.200806 name = ''febo'' memory = ''256'' vif = [ ''mac=00:16:3e:1b:04:c9, ip=10.1.0.70'' ] # Install disk = [ ''file:/root/os200805.iso,6:cdrom,r'', ''phy:vm1/febo,0,w'' ] kernel = ''/etc/xen/os1/unix'' ramdisk = ''/etc/xen/os1/x86.microroot'' extra = ''/platform/i86xpv/kernel/unix -B console=ttya,livemode=text'' on_shutdown = ''destroy'' on_reboot = ''destroy'' on_crash = ''destroy'' Booting I got a panic message: sirius:/etc/xen# xm create -c febo.cfg.200806 Using config file "./febo.cfg.200806". v3.2-1 chgset ''unavailable'' Started domain febo SunOS Release 5.11 Version snv_86 32-bit Copyright 1983-2008 Sun Microsystems, Inc. All rights reserved. Use is subject to license terms. panic[cpu0]/thread=f5028f60sirius:/etc/xen# In [1] I noticed the following difference in the configuration file: extra = "/platform/i86xpv/kernel/unix -kd - nowin -B install_media=cdrom" Introducing ''gnttab_init+0xce/W 403'' during the Solaris bootstrapping, I got this stack trace: sirius:/etc/xen# xm create -c febo.cfg.200806 Using config file "./febo.cfg.200806". Started domain febo Loading kmdb... Welcome to kmdb Loaded modules: [ unix krtld genunix ] [0]> gnttab_init+0xce/W 403 gnttab_init+0xce: 0x6ac472 = 0x403 [0]> :c v3.2-1 chgset ''unavailable'' SunOS Release 5.11 Version snv_86 32-bit Copyright 1983-2008 Sun Microsystems, Inc. All rights reserved. Use is subject to license terms. panic[cpu0]/thread=f5028f60: BAD TRAP: type=e (#pf Page fault) rp=f5053654 addr=d4800570 #pf Page fault Bad kernel fault at addr=0xd4800570 pid=0, pc=0xf4c70916, sp=0x0, eflags=0x10002 cr0: 80050033<pg,wp,ne,et,mp,pe> cr4: 660<xmme,fxsr,mce,pae> cr2: d4800570 gs: f4c101b1 fs: 1bd10000 es: 80000161 ds: d4800161 edi: 1bc673 esi: d4800570 ebp: f5053700 esp: f505368c ebx: 1bc673 edx: 0 ecx: 80000000 eax: 0 trp: e err: 3 eip: f4c70916 cs: 159 efl: 10002 usp: 0 ss: 0 f50535ac unix:die+ed (e, f5053654, d48005) f5053640 unix:trap+12b9 (f5053654, d4800570,) f5053654 unix:_cmntrap+aa (f4c101b1, 1bd10000,) f5053700 unix:atomic_cas_64+16 (d4611ea0, ae, 1bc67) f505375c unix:hati_pte_map+143 (d4611ea0, ae, 0, 1b) f50537a0 unix:hati_load_common+12f (d4460f68, d52ae000,) f5053804 unix:hat_devload+10b (d4460f68, d52ae000,) f5053858 unix:gnttab_init+bb (f4c4feae, 3, f50538) f5053868 unix:startup_modules+fb (f5053880, f4d3d812,) f5053870 unix:startup+35 (f4c00010, f504f688,) f5053880 genunix:main+1e () panic: entering debugger (no dump device, continue to reboot) Loaded modules: [ specfs ] kmdb: target stopped at: kmdb_enter+0xa: pushl %eax [0]> Would I have to download the kernel packages mentioned in [1]? Regards, Daniel [1] http://lxer.com/module/newswire/view/103407/index.html -- Daniel Bareiro - System Administrator Fingerprint: BFB3 08D6 B4D1 31B2 72B9 29CE 6696 BF1B 14E6 1D37 Powered by Debian GNU/Linux Lenny - Linux user #188.598
Daniel, All written bellow works on Ubuntu 8.04, but not on Debian Etch or any other Debian release. If you want to continue on Debian then you are on your own, i cannot help Patched Ubuntu 8.04 kernel is already in repositories. View:- 1. http://www.howtoforge.com/high-performance-xen-on-ubuntu-8.04-amd64 It is 2.6.24-17-xen View also procedure at :- http://www.opensolaris.org/jive/thread.jspa?threadID=62436&tstart=0 # echo deb http://ftp.iinet.net.au/pub/ubuntu hardy-updates main universe multiverse >>/etc/apt/sources.list # apt-get install grub linux-image-xen If you will follow it without "chroot /mnt" (which installs xen on the image FS) kernel 2.6.24-18-xen will be installed. ******************************************************************************** One more notice, if your setup is Desktop (not Server) you might have problems with Network-Manager. I am not sure it has been fixed. Your start (on Desktop) should follow:- http://deshantm.livejournal.com/29959.html ******************************************************************************** sudo apt-get remove network-manager sudo gedit /etc/network/interfaces auto lo iface lo inet loopback auto eth0 iface eth0 inet dhcp sudo /etc/init.d/network restart sudo apt-get install ubuntu-xen-server ************************************************* Todd''s "HowTo" is relatively old, so you have to skip three next commands in his instruction, applying patch to the kernel. Thanks. -b This message posted from opensolaris.org
> El domingo 08 de junio del 2008 a las 08:23:28, > Boris Derzhavets escribió: > > > Daniel, > > Hi Boris. > > > You are right. However, install procedure was done > by myself at Xen > > 3.2.1 Ubuntu 8.04, CentOS 5.1 Dom0 ( thanks to > Jurgen Keil and Mark > > Jonson) :- > http://lxer.com/module/newswire/view/103407/index.html > > > http://lxer.com/module/newswire/view/103407/index.html > and at Xen 3.1 > > F8 Dom0 with no problems (no kernel patch needed) > > I try installing xen-utils-3.2-1 and > xen-hypervisor-3.2-1-i386 packages > from backports repositories. These have Xen 3.2.0-3 > version, according > to I can see. This is domU's config file: > > sirius:/etc/xen# cat febo.cfg.200806 > name = 'febo' > memory = '256' > vif = [ 'mac=00:16:3e:1b:04:c9, ip=10.1.0.70' ] > > # Install > disk = [ 'file:/root/os200805.iso,6:cdrom,r', > 'phy:vm1/febo,0,w' ] > kernel = '/etc/xen/os1/unix' > ramdisk = '/etc/xen/os1/x86.microroot' > extra = '/platform/i86xpv/kernel/unix -B > console=ttya,livemode=text' > on_shutdown = 'destroy' > on_reboot = 'destroy' > on_crash = 'destroy' > > Booting I got a panic message: > > sirius:/etc/xen# xm create -c febo.cfg.200806 > Using config file "./febo.cfg.200806". > v3.2-1 chgset 'unavailable' > Started domain febo > SunOS Release 5.11 Version snv_86 32-bit > Copyright 1983-2008 Sun Microsystems, Inc. All > rights reserved. > Use is subject to license terms. > > panic[cpu0]/thread=f5028f60sirius:/etc/xen# > > In [1] I noticed the following difference in the > configuration file: > > extra = "/platform/i86xpv/kernel/unix -kd - nowin -B > install_media=cdrom" > > Introducing 'gnttab_init+0xce/W 403' during the > Solaris bootstrapping, I**************************************************************** This is for 64-bit version , not 32-bit Regarding 32-bit view:- http://www.opensolaris.org/jive/thread.jspa?threadID=55632&tstart=30 Per Jurgen Keil:- And the workaround would be: # xm create /tmp/solaris-exp -c Using config file "/tmp/solaris-exp". Loading kmdb... Started domain solaris-exp Welcome to kmdb Loaded modules: [ unix krtld genunix ] [0]> gnttab_init+0xa2:b ,5000:c v3.2.0 chgset 'unavailable' SunOS Release 5.11 Version snv_85 32-bit Copyright 1983-2008 Sun Microsystems, Inc. All rights reserved. Use is subject to license terms. Loaded modules: [ specfs ] t0stack+0x1f1c: 0x3 = 0x403 kmdb: stop at gnttab_init+0xa2 kmdb: target stopped at: gnttab_init+0xa2: pushl %eax t0stack+0x1f1c: 0x3 = 0x403 kmdb: stop at gnttab_init+0xa2 kmdb: target stopped at: gnttab_init+0xa2: pushl %eax ********************************************************** It will allow you to startup. However, you will have to patch kernel as file on 32-bit system. Jurgen didn't provide any instructions for that step. **********************************************************> got this stack trace: > > sirius:/etc/xen# xm create -c febo.cfg.200806 > Using config file "./febo.cfg.200806". > Started domain febo > Loading kmdb... > aded modules: [ unix krtld genunix ] > [0]> gnttab_init+0xce/W 403 > gnttab_init+0xce: 0x6ac472 > 0x403 > c > v3.2-1 chgset 'unavailable' > SunOS Release 5.11 Version snv_86 32-bit > Copyright 1983-2008 Sun Microsystems, Inc. All > rights reserved. > Use is subject to license terms. > > panic[cpu0]/thread=f5028f60: BAD TRAP: type=e (#pf > Page fault) rp=f5053654 addr=d4800570 > > #pf Page fault > Bad kernel fault at addr=0xd4800570 > pid=0, pc=0xf4c70916, sp=0x0, eflags=0x10002 > cr0: 80050033<pg,wp,ne,et,mp,pe> cr4: > 660<xmme,fxsr,mce,pae> > cr2: d4800570 > gs: f4c101b1 fs: 1bd10000 es: 80000161 > ds: d4800161 > edi: 1bc673 esi: d4800570 ebp: f5053700 esp: > f505368c > ebx: 1bc673 edx: 0 ecx: 80000000 eax: > 0 > trp: e err: 3 eip: f4c70916 cs: > 159 > efl: 10002 usp: 0 ss: 0 > 0535ac unix:die+ed (e, f5053654, d48005) > f5053640 unix:trap+12b9 (f5053654, d4800570,) > f5053654 unix:_cmntrap+aa (f4c101b1, 1bd10000,) > f5053700 unix:atomic_cas_64+16 (d4611ea0, ae, 1bc67) > f505375c unix:hati_pte_map+143 (d4611ea0, ae, 0, 1b) > f50537a0 unix:hati_load_common+12f (d4460f68, > d52ae000,) > f5053804 unix:hat_devload+10b (d4460f68, d52ae000,) > f5053858 unix:gnttab_init+bb (f4c4feae, 3, f50538) > f5053868 unix:startup_modules+fb (f5053880, > f4d3d812,) > f5053870 unix:startup+35 (f4c00010, f504f688,) > f5053880 genunix:main+1e () > > panic: entering debugger (no dump device, continue to > reboot) > Loaded modules: [ specfs ] > kmdb: target stopped at: > kmdb_enter+0xa: pushl %eax > [0]> > > Would I have to download the kernel packages > mentioned in [1]? > > Regards, > Daniel > > [1] > http://lxer.com/module/newswire/view/103407/index.html > -- > Daniel Bareiro - System Administrator > Fingerprint: BFB3 08D6 B4D1 31B2 72B9 29CE 6696 BF1B > 14E6 1D37 > Powered by Debian GNU/Linux Lenny - Linux user > #188.598 > > _______________________________________________ > xen-discuss mailing list > xen-discuss@opensolaris.orgThis message posted from opensolaris.org _______________________________________________ xen-discuss mailing list xen-discuss@opensolaris.org
> El domingo 08 de junio del 2008 a las 08:23:28, > Boris Derzhavets escribió: > > > Daniel, > > Hi Boris. > > > You are right. However, install procedure was done > by myself at Xen > > 3.2.1 Ubuntu 8.04, CentOS 5.1 Dom0 ( thanks to > Jurgen Keil and Mark > > Jonson) :- > http://lxer.com/module/newswire/view/103407/index.html > > > http://lxer.com/module/newswire/view/103407/index.html > and at Xen 3.1 > > F8 Dom0 with no problems (no kernel patch needed) > > I try installing xen-utils-3.2-1 and > xen-hypervisor-3.2-1-i386 packages > from backports repositories. These have Xen 3.2.0-3******************************************************************** If you want to stay on Debian Etch, then there is a good "HowTo" http://www.howtoforge.com/debian_etch_xen_3.1 for 32-bit version. OpenSolaris 2008.05 is based on SNV86 . This Nevada build has problem with Xen 3.2 Linux Dom0. It may be resolved, but what for ? ( if your major target is OS200805 DomU ) Once again , if you started "apt-getting" on Debian then getting Xen 3.1 properly installed would work fine for OpenSolaris 2008.05 *******************************************************************> version, according > to I can see. This is domU's config file: > > sirius:/etc/xen# cat febo.cfg.200806 > name = 'febo' > memory = '256' > vif = [ 'mac=00:16:3e:1b:04:c9, ip=10.1.0.70' ] > > # Install > disk = [ 'file:/root/os200805.iso,6:cdrom,r', > 'phy:vm1/febo,0,w' ] > kernel = '/etc/xen/os1/unix' > ramdisk = '/etc/xen/os1/x86.microroot' > extra = '/platform/i86xpv/kernel/unix -B > console=ttya,livemode=text' > on_shutdown = 'destroy' > on_reboot = 'destroy' > on_crash = 'destroy' > > Booting I got a panic message: > > sirius:/etc/xen# xm create -c febo.cfg.200806 > Using config file "./febo.cfg.200806". > v3.2-1 chgset 'unavailable' > Started domain febo > SunOS Release 5.11 Version snv_86 32-bit > Copyright 1983-2008 Sun Microsystems, Inc. All > rights reserved. > Use is subject to license terms. > > panic[cpu0]/thread=f5028f60sirius:/etc/xen# > > In [1] I noticed the following difference in the > configuration file: > > extra = "/platform/i86xpv/kernel/unix -kd - nowin -B > install_media=cdrom" > > Introducing 'gnttab_init+0xce/W 403' during the > Solaris bootstrapping, I > got this stack trace: > > sirius:/etc/xen# xm create -c febo.cfg.200806 > Using config file "./febo.cfg.200806". > Started domain febo > Loading kmdb... > aded modules: [ unix krtld genunix ] > [0]> gnttab_init+0xce/W 403 > gnttab_init+0xce: 0x6ac472 > 0x403 > c > v3.2-1 chgset 'unavailable' > SunOS Release 5.11 Version snv_86 32-bit > Copyright 1983-2008 Sun Microsystems, Inc. All > rights reserved. > Use is subject to license terms. > > panic[cpu0]/thread=f5028f60: BAD TRAP: type=e (#pf > Page fault) rp=f5053654 addr=d4800570 > > #pf Page fault > Bad kernel fault at addr=0xd4800570 > pid=0, pc=0xf4c70916, sp=0x0, eflags=0x10002 > cr0: 80050033<pg,wp,ne,et,mp,pe> cr4: > 660<xmme,fxsr,mce,pae> > cr2: d4800570 > gs: f4c101b1 fs: 1bd10000 es: 80000161 > ds: d4800161 > edi: 1bc673 esi: d4800570 ebp: f5053700 esp: > f505368c > ebx: 1bc673 edx: 0 ecx: 80000000 eax: > 0 > trp: e err: 3 eip: f4c70916 cs: > 159 > efl: 10002 usp: 0 ss: 0 > 0535ac unix:die+ed (e, f5053654, d48005) > f5053640 unix:trap+12b9 (f5053654, d4800570,) > f5053654 unix:_cmntrap+aa (f4c101b1, 1bd10000,) > f5053700 unix:atomic_cas_64+16 (d4611ea0, ae, 1bc67) > f505375c unix:hati_pte_map+143 (d4611ea0, ae, 0, 1b) > f50537a0 unix:hati_load_common+12f (d4460f68, > d52ae000,) > f5053804 unix:hat_devload+10b (d4460f68, d52ae000,) > f5053858 unix:gnttab_init+bb (f4c4feae, 3, f50538) > f5053868 unix:startup_modules+fb (f5053880, > f4d3d812,) > f5053870 unix:startup+35 (f4c00010, f504f688,) > f5053880 genunix:main+1e () > > panic: entering debugger (no dump device, continue to > reboot) > Loaded modules: [ specfs ] > kmdb: target stopped at: > kmdb_enter+0xa: pushl %eax > [0]> > > Would I have to download the kernel packages > mentioned in [1]? > > Regards, > Daniel > > [1] > http://lxer.com/module/newswire/view/103407/index.html > -- > Daniel Bareiro - System Administrator > Fingerprint: BFB3 08D6 B4D1 31B2 72B9 29CE 6696 BF1B > 14E6 1D37 > Powered by Debian GNU/Linux Lenny - Linux user > #188.598 > > _______________________________________________ > xen-discuss mailing list > xen-discuss@opensolaris.orgThis message posted from opensolaris.org _______________________________________________ xen-discuss mailing list xen-discuss@opensolaris.org
Boris Derzhavets wrote:> This is for 64-bit version , not 32-bit > Regarding 32-bit view:- > http://www.opensolaris.org/jive/thread.jspa?threadID=55632&tstart=30Correct, the 64-bit kmdb patch procedure doesn''t work when you boot a 32-bit domU. Daniel must use the 32-bit vesion of the kmdb patch...> Per Jurgen Keil:- > And the workaround would be: > [ ... kmdb patch for booting a 32-bit PV domU on xen-3.2 ... ]> ********************************************************** > It will allow you to startup. > However, you will have to patch kernel as file > on 32-bit system. Jurgen didn''t provide any > instructions for that step. > **********************************************************It''s a bit difficult to patch that into the 32-bit kernel module file, because the new 0x403 value needs more space than we have available. (There''s no such problem with the 64-bit kernel) But: the same 32-bit kmdb patch can be used to boot the installed domU, and as a next step you can bfu the whole SXCE domU to something newer than build 87; after the bfu upgrade of the system the kmdb patch wouldn''t be necessary any more. (or just start SXCE domU installation with a current build). And it seems that for OpenSolaris 2008.05 you can now update to build 90, see this thread (and note the special grub mbr installation instructions at the end): http://www.opensolaris.org/jive/thread.jspa?threadID=62982&tstart=0 This message posted from opensolaris.org
Sorry, for asking same question second time. How to bfu SNV87 installed on bare metal (with 30KB/sec ADSL Internet) to SNV91 ? I would be happy to go just once through step-by-step described procedure, which would be understandable for person not doing Solaris 24x7. The link suggested to me first time when i asked for instructions is above my compehension This message posted from opensolaris.org
> Sorry, for asking same question second time. > How to bfu SNV87 installed on bare metal (with > 30KB/sec ADSL Internet) to SNV91 ?1. download and install the "ON Specific Build Tools" package (for the x86 platform) (~ a 1 Mbyte download) # cd /usr/tmp # /usr/sfw/bin/wget http://dlc.sun.com/osol/on/downloads/b91/SUNWonbld.i386.tar.bz2 # bunzip2< SUNWonbld.i386.tar.bz2 | tar xf - # pkgadd -d onbld SUNWonbld (answer ''y'' to the question asking for permission to run scripts that will be run with super-user priviledges) 2. download an unpack the "ON BFU Archives (non-debug)" (for the x86 platform) (~ a 110 MByte download) # /usr/sfw/bin/wget http://dlc.sun.com/osol/on/downloads/b91/on-bfu-nightly-osol-nd.i386.tar.bz2 # bunzip2 < on-bfu-nightly-osol-nd.i386.tar.bz2 | tar xf - 3. upgrade the system using "bfu": (I''m assuming you''re using the bourne or bash shell) First we need some evironment variables: # PATH=/opt/onbld/bin:$PATH # FASTFS=/opt/onbld/bin/`uname -p`/fastfs # BFULD=/opt/onbld/bin/`uname -p`/bfuld # GZIPBIN=/usr/bin/gzip # ACR=/opt/onbld/bin/acr # export PATH FASTFS BFULD GZIPBIN ACR Now upgrade the system and install the new onnv bits unpacked in step 2. # bfu /usr/tmp/archives-nightly-osol-nd/i386 Run the "acr" command (this automatically merges your local configuration file changes with the new bits that have just been installed) # acr Reboot the upgraded system: # reboot System should now boot using a snv_91 kernel. This message posted from opensolaris.org
Afaik 91 is not out, 90 is current for external. Download onbld and ON bfu archive (non-debug) (Total: ~110mb) As root: Use pkgadd to install current onbld tools Set environment variables - FASTFS="/opt/onbld/bin/i386/fastfs"; export FASTFS BFULD="/opt/onbld/bin/bfuld"; export BFULD GZIPBIN="/usr/bin/gzip"; export GZIPBIN Extract bfu archive which extracts "archives" directory /opt/onbld/bin/bfu /path/to/archives/i386/nightly Run ''acr'' to resolve conflicts (In onbld) Update JDS, XWIN (X.org), SFW packages after reboot into new environment. Total ~1gb. B91 using lu would be better but it''s not there, so you have to upgrade bits for now. Not sure about how this will affect Xen though, since the tools aren''t exactly part of ON. James On Jun 9, 2008, at 11:11 AM, Boris Derzhavets wrote:> Sorry, for asking same question second time. > How to bfu SNV87 installed on bare metal (with 30KB/sec ADSL Internet) > to SNV91 ? > I would be happy to go just once through step-by-step described > procedure, > which would be understandable for person not doing Solaris 24x7. > The link suggested to me first time when i asked for instructions is > above > my compehension > > > This message posted from opensolaris.org > _______________________________________________ > xen-discuss mailing list > xen-discuss@opensolaris.org
Oh sorry about previous reply I didn''t see you were on 2008.5... I think that it''s an svr4 package. James On Jun 9, 2008, at 11:11 AM, Boris Derzhavets wrote:> Sorry, for asking same question second time. > How to bfu SNV87 installed on bare metal (with 30KB/sec ADSL Internet) > to SNV91 ? > I would be happy to go just once through step-by-step described > procedure, > which would be understandable for person not doing Solaris 24x7. > The link suggested to me first time when i asked for instructions is > above > my compehension > > > This message posted from opensolaris.org > _______________________________________________ > xen-discuss mailing list > xen-discuss@opensolaris.org
I never was on 2008.05 as Dom0. Currently running Solaris xVM ( SNV87) on bare metal and SNV89(90) as DomUs at Xen 3.2 F8 Dom0 This message posted from opensolaris.org
Thanks a lot. Instructions look pretty clear. Would xVM be automatically upgraded along with this procedure ?
Thank you. -b
Thank you once again. I''ve bfu''ed SNV_87 to SNV_91 in about 1 hr. xm info reports :- xen_changeset : Mon Jun 11 14:56:50 2007 +0100 15634:1a366250feeb So, i guess xVM updated as well. As soon as SNV91 will become available for download, i''ll just bfu SNV90 HVM DomU to SNV91 and run fork_100 to make sure the bug is fixed in the release. I''ve also understood your suggestion to bfu OS200805 (32-bit) HVM DomU to SNV90 kernel and eliminate issue with kernel patch. It won''t take long time (1-2 hr).
> Thanks a lot. Instructions look pretty clear. > Would xVM be automatically upgraded along with this > procedure ?Only those parts of xVM that are part of the ON "OS/Net consolidation". This includes the dom0 / domU PV kernel bits, the PV network and disk drivers, and the HVM PV drivers. The rest of xVM (the hypervisor, xend, xenconsole, the qemu device emulation for HVM domU, virt-install ...) is part of a different set of sources: it''s the "xVM Source" download on http://dlc.sun.com/osol/on/downloads/b91/ There are no pre-build binary packages available for that. Building from source is something like this: $ /usr/sfw/bin/wget http://dlc.sun.com/osol/on/downloads/b91/xvm-src.tar.bz2 $ bunzip2 < xvm-src.tar.bz2 | tar xf - $ EMAIL=your_email@foo.bar.com $ XVM_WS=`pwd` $ SOLARIS_BUILD_TOOLS=/opt $ LD_OPTIONS=-L/lib $ export EMAIL XVM_WS SOLARIS_BUILD_TOOLS LD_OPTIONS $ sunos.hg/bin/build-all nondebug When all goes well, you should find 9 installable packages for xVM in the "packages-nondebug" subdirectory. To install them: $ su # svcadm disable xvm/domains xvm/console xvm/xend xvm/store # pkgrm SUNWvirtinst SUNWurlgrabber SUNWlibvirt SUNWxvmh SUNWxvmhvm SUNWxvmu SUNWxvmr SUNWxvmdomu SUNWxvmdomr # pkgadd -d packages-nondebug SUNWxvmdomr SUNWxvmdomu SUNWxvmr SUNWxvmu SUNWxvmhvm SUNWxvmh SUNWlibvirt SUNWurlgrabber SUNWvirtinst # svcadm enable xvm/store xvm/xend xvm/console xvm/domains # reboot This message posted from opensolaris.org
> I''ve bfu''ed SNV_87 to SNV_91 in about 1 hr. > xm info reports > xen_changeset : Mon Jun 11 14:56:50 2007 +0100 15634:1a366250feeb > So, i guess xVM updated as well.No, bfu doesn''t update the hypervisor. Note that the latest "xen changeset" that was applied to the hypervisor is from 2007, exactly one year ago. Which surprises me a bit, because SNV_87 is only three month old. Is /boot/xen.gz /boot/amd64/xen.gz the orignal hypervisor that was included with SNV_87, or did you manually replace it with some other version?> As soon as SNV91 will become available > for download, i''ll just bfu SNV90 HVM DomU to > SNV91 and run fork_100 to make sure the bug is > fixed in the release.You have bfu HVM / PV SXCE domUs just like you did bfu your dom0. This message posted from opensolaris.org
On Tue, Jun 10, 2008 at 03:25:44AM -0700, J??rgen Keil wrote:> > I''ve bfu''ed SNV_87 to SNV_91 in about 1 hr. > > xm info reports > > xen_changeset : Mon Jun 11 14:56:50 2007 +0100 15634:1a366250feeb > > So, i guess xVM updated as well. > > No, bfu doesn''t update the hypervisor. Note that the latest "xen changeset" > that was applied to the hypervisor is from 2007, exactly one year ago.This changeset doesn''t reflect the actual changes in the binary (annoyingly) regards john
> > I''ve bfu''ed SNV_87 to SNV_91 in about 1 hr. > > xm info reports > > xen_changeset : Mon Jun 11 14:56:50 2007 > +0100 15634:1a366250feeb > > So, i guess xVM updated as well. > > No, bfu doesn''t update the hypervisor. Note that the > latest "xen changeset" > that was applied to the hypervisor is from 2007, > exactly one year ago. > > Which surprises me a bit, because SNV_87 is only > three month old. > > Is /boot/xen.gz /boot/amd64/xen.gz the orignal > hypervisor that was > included with SNV_87, or did you manually replace it > with some other > version?*********************************************************** I didn''t touch anything after install form SNV87 DVD ************************************************************> > > As soon as SNV91 will become available > > for download, i''ll just bfu SNV90 HVM DomU to > > SNV91 and run fork_100 to make sure the bug is > > fixed in the release. > > You can bfu HVM / PV SXCE domUs just like you did > bfu your dom0. > > Message was edited by: > jkeilThis message posted from opensolaris.org
Attempt to reproduce:- bash-3.2# cd /usr/tmp bash-3.2# ls -ld *.hg drwxr-xr-x 4 300 8190 512 Jun 4 21:52 libvirt.hg drwxr-xr-x 6 300 8190 512 Jun 4 21:52 sunos.hg drwxr-xr-x 3 300 8190 512 Jun 4 21:52 urlgrabber.hg drwxr-xr-x 5 300 8190 512 Jun 4 21:55 virtinst.hg drwxr-xr-x 12 300 8190 512 Jun 4 21:54 xen.hg bash-3.2# ls -l set* -rwxr--r-- 1 root root 186 Jun 10 09:09 setBFUenv -rwxr--r-- 1 root root 135 Jun 10 15:16 setBFUXenEnv bash-3.2# . ./setBFUXenEnv bash-3.2# env|grep EMAIL EMAIL=bderzhavets@yahoo.com bash-3.2# env|grep XVM_WS XVM_WS=/usr/tmp bash-3.2# env|grep SOLARIS_BUILD_TOOLS SOLARIS_BUILD_TOOLS=/opt bash-3.2# env|grep LD_OPTIONS LD_OPTIONS=-L/lib bash-3.2# sunos.hg/bin/build-all nondebug /usr/tmp/xen.hg does not seem to be under MQ management This message posted from opensolaris.org
> Attempt to reproduce:...> bash-3.2# sunos.hg/bin/build-all nondebug > /usr/tmp/xen.hg does not seem to be under MQ managementAhh, you also have to enable the "mercurial queue" (MQ) extension In your $HOME directory, create a file like this: % cat $HOME/.hgrc [extensions] hgext.mq[diff] git=True
Looks like Sun Pro C is required:- ************************************************* ************************************************* * WARNING: Package ''latex'' is required * to build Xen documentation ************************************************* ************************************************* ginstall -d -m0755 -p /usr/tmp/proto/check ginstall -m0644 -p ./COPYING /usr/tmp/proto ginstall -m0644 -p ./README /usr/tmp/proto ginstall -m0755 -p ./install.sh /usr/tmp/proto ginstall -m0755 -p tools/check/chk tools/check/check_* /usr/tmp/proto/check gmake[1]: Leaving directory `/var/tmp/xen.hg'' ---- fixing for vendor-packages ---- 7552 blocks ---- preparing SFW gate /usr/tmp/libvirt.hg ---- ---- building libvirt ---- ---- switching to 64-bit build ---- checking build system type... i386-pc-solaris2.11 checking host system type... i386-pc-solaris2.11 Beginning configuration for readline-5.2 for i386-pc-solaris2.11 checking whether /usr/sfw/bin/gmake sets $(MAKE)... yes checking for gcc... /opt/SUNWspro/SS11/bin/cc checking for C compiler default output file name... configure: error: C compiler cannot create executables See `config.log'' for more details. --- On Tue, 6/10/08, Juergen Keil <jk@tools.de> wrote: From: Juergen Keil <jk@tools.de> Subject: Re: [xen-discuss] OpenSolaris 2008.05 domU on Debian dom0 To: bderzhavets@yahoo.com Cc: xen-discuss@opensolaris.org Date: Tuesday, June 10, 2008, 12:14 PM > Attempt to reproduce: ... > bash-3.2# sunos.hg/bin/build-all nondebug > /usr/tmp/xen.hg does not seem to be under MQ management Ahh, you also have to enable the "mercurial queue" (MQ) extension In your $HOME directory, create a file like this: % cat $HOME/.hgrc [extensions] hgext.mq[diff] git=True
El lunes 09 de junio del 2008 a las 21:34:34, Boris Derzhavets escribió:> I never was on 2008.05 as Dom0. Currently running Solaris xVM ( SNV87) > on bare metal and SNV89(90) as DomUs at Xen 3.2 F8 Dom0Mmmmm... I am losing a little in this thread :-S What are the differences between 200805 and SNV*? On the other hand, also I tried to dowload the ISO image for DVD of Solaris 10 from the site [1] of Sun for to install the operating system in a domU Xen with Debian GNU/Linux dom0 and to also see if with that distribution of OpenSolaris I have more luck, but seems that there is a problem with the URL (sun-10-u5-ga-x86-dvd.iso) or the image, every time I already tried download the file was empty. Regards, Daniel [1] http://www.sun.com/software/solaris/get.jsp -- Daniel Bareiro - System Administrator Fingerprint: BFB3 08D6 B4D1 31B2 72B9 29CE 6696 BF1B 14E6 1D37 Powered by Debian GNU/Linux Lenny - Linux user #188.598
Made a link and trying again:-
bash-3.2# ls -l /opt/SUNWspro/SS11/bin/cc
lrwxrwxrwx 1 root
root
16 Jun 10 17:02 /opt/SUNWspro/SS11/bin/cc -> /usr/sfw/bin/gcc
--- On Tue, 6/10/08, Boris Derzhavets <bderzhavets@yahoo.com>
wrote:
From: Boris Derzhavets <bderzhavets@yahoo.com>
Subject: Re: [xen-discuss] OpenSolaris 2008.05 domU on Debian dom0
To: "Juergen Keil" <jk@tools.de>
Cc: xen-discuss@opensolaris.org
Date: Tuesday, June 10, 2008, 12:51 PM
Looks like Sun Pro C is required:-
*************************************************
*************************************************
* WARNING: Package ''latex'' is required
*
to build Xen documentation
*************************************************
*************************************************
ginstall -d -m0755 -p /usr/tmp/proto/check
ginstall -m0644 -p ./COPYING /usr/tmp/proto
ginstall -m0644 -p ./README /usr/tmp/proto
ginstall -m0755 -p ./install.sh /usr/tmp/proto
ginstall -m0755 -p tools/check/chk tools/check/check_* /usr/tmp/proto/check
gmake[1]: Leaving directory `/var/tmp/xen.hg''
---- fixing for vendor-packages ----
7552
blocks
---- preparing SFW gate /usr/tmp/libvirt.hg ----
---- building libvirt ----
---- switching to 64-bit build ----
checking build system type... i386-pc-solaris2.11
checking host system type... i386-pc-solaris2.11
Beginning configuration for readline-5.2 for i386-pc-solaris2.11
checking whether /usr/sfw/bin/gmake sets $(MAKE)... yes
checking for gcc... /opt/SUNWspro/SS11/bin/cc
checking for C compiler default output file name... configure: error: C compiler
cannot create executables
See `config.log'' for more details.
--- On Tue, 6/10/08, Juergen Keil <jk@tools.de> wrote:
From: Juergen Keil <jk@tools.de>
Subject: Re: [xen-discuss] OpenSolaris 2008.05 domU on Debian dom0
To: bderzhavets@yahoo.com
Cc: xen-discuss@opensolaris.org
Date: Tuesday, June 10, 2008, 12:14
PM
> Attempt to reproduce:
...
> bash-3.2# sunos.hg/bin/build-all nondebug
> /usr/tmp/xen.hg does not seem to be under MQ management
Ahh, you also have to enable the "mercurial queue" (MQ) extension
In your $HOME directory, create a file like this:
% cat $HOME/.hgrc
[extensions]
hgext.mq[diff]
git=True
_______________________________________________
xen-discuss mailing list
xen-discuss@opensolaris.org
> Looks like Sun Pro C is required:...> ---- building libvirt...> checking whether /usr/sfw/bin/gmake sets $(MAKE)... yes > checking for gcc... /opt/SUNWspro/SS11/bin/cc > checking for C compiler default output file name... > configure: error: C compiler cannot create executablesI do remember... Not only is Sun Pro C required, you also must have that SS11 directory inside SUNWspro. I did have Sun Pro C installed (it required for building opensolaris from source), but the SS11 subdirectory was missing. I just created a symlink /opt/SUNWspro/SS11 pointing to /opt/SUNWspro # ln -s . /opt/SUNWspro/SS11 # ls -l /opt/SUNWspro/SS11 lrwxrwxrwx 1 root root 1 Mar 17 10:40 /opt/SUNWspro/SS11 -> . # % ls -l /opt/SUNWspro/SS11/bin/cc lrwxrwxrwx 1 root other 14 Jun 22 2007 /opt/SUNWspro/SS11/bin/cc -> ../prod/bin/cc This message posted from opensolaris.org
Build failed . Config.log is attached
--- On Tue, 6/10/08, Boris Derzhavets <bderzhavets@yahoo.com>
wrote:
From: Boris Derzhavets <bderzhavets@yahoo.com>
Subject: Re: [xen-discuss] OpenSolaris 2008.05 domU on Debian dom0
To: "Juergen Keil" <jk@tools.de>
Cc: xen-discuss@opensolaris.org
Date: Tuesday, June 10, 2008, 1:08 PM
Made a link and trying again:-
bash-3.2# ls -l /opt/SUNWspro/SS11/bin/cc
lrwxrwxrwx 1 root
root
16 Jun 10 17:02 /opt/SUNWspro/SS11/bin/cc -> /usr/sfw/bin/gcc
--- On Tue, 6/10/08, Boris Derzhavets <bderzhavets@yahoo.com>
wrote:
From: Boris Derzhavets <bderzhavets@yahoo.com>
Subject: Re: [xen-discuss] OpenSolaris 2008.05 domU on Debian dom0
To: "Juergen Keil" <jk@tools.de>
Cc: xen-discuss@opensolaris.org
Date: Tuesday, June 10, 2008, 12:51 PM
Looks like Sun Pro C is required:-
*************************************************
*************************************************
* WARNING: Package ''latex'' is required
*
to build Xen documentation
*************************************************
*************************************************
ginstall -d -m0755 -p /usr/tmp/proto/check
ginstall -m0644 -p ./COPYING /usr/tmp/proto
ginstall -m0644 -p ./README /usr/tmp/proto
ginstall -m0755 -p ./install.sh /usr/tmp/proto
ginstall -m0755
-p tools/check/chk tools/check/check_* /usr/tmp/proto/check
gmake[1]: Leaving directory `/var/tmp/xen.hg''
---- fixing for vendor-packages ----
7552
blocks
---- preparing SFW gate /usr/tmp/libvirt.hg ----
---- building libvirt ----
---- switching to 64-bit build ----
checking build system type... i386-pc-solaris2.11
checking host system type... i386-pc-solaris2.11
Beginning configuration for readline-5.2 for i386-pc-solaris2.11
checking whether /usr/sfw/bin/gmake sets $(MAKE)... yes
checking for gcc... /opt/SUNWspro/SS11/bin/cc
checking for C compiler default output file name... configure: error: C compiler
cannot create executables
See `config.log'' for more details.
--- On Tue, 6/10/08, Juergen Keil <jk@tools.de> wrote:
From: Juergen Keil <jk@tools.de>
Subject: Re: [xen-discuss] OpenSolaris 2008.05 domU on Debian dom0
To: bderzhavets@yahoo.com
Cc: xen-discuss@opensolaris.org
Date: Tuesday, June 10, 2008, 12:14
PM
> Attempt to reproduce:
...
> bash-3.2# sunos.hg/bin/build-all nondebug
> /usr/tmp/xen.hg does not seem to be under MQ management
Ahh, you also have to enable the "mercurial queue" (MQ) extension
In your $HOME directory, create a file like this:
% cat $HOME/.hgrc
[extensions]
hgext.mq[diff]
git=True
_______________________________________________
xen-discuss mailing list
xen-discuss@opensolaris.org
_______________________________________________
xen-discuss mailing list
xen-discuss@opensolaris.org
> Made a link and trying again:- > > bash-3.2# ls -l /opt/SUNWspro/SS11/bin/cc > lrwxrwxrwx 1 root root 16 Jun 10 17:02 /opt/SUNWspro/SS11/bin/cc ->/usr/sfw/bin/gcc Apparently, that trick didn''t work, ...> configure:1844: /opt/SUNWspro/SS11/bin/cc --version </dev/null >&5 > cc (GCC) 3.4.3 (csl-sol210-3_4-20050802) > Copyright (C) 2004 Free Software Foundation, Inc. > This is free software; see the source for copying conditions. There is NO > warranty; not even for MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. > ... > cc: unrecognized option `-Xa'' > cc: unrecognized option `-KPIC'' > cc: language arch=generic64 not recognized... it is passing SunPro C compiler options to GNU C; and GNU C gets confused. Try to install the Sun Pro C compiler, if you havn''t done so. You should find it on the SXCE installation DVD, in the subdirectory "DeveloperTools". DeveloperTools/README has installation instructions; apparently all you have to do is run DeveloperTools/install_devtools.sh as user root (and that needs another 500MB of free space in /opt) And in case /opt/SUNWspro/SS11 isn''t present after installation, install a symlink /opt/SUNWspro/SS11 -> .
> Mmmmm... I am losing a little in this thread :-S What are the< differences between 200805 and SNV*? OpenSolaris 2008.05 (http://www.opensolaris.com/) - uses a Live-CD, from there you can start the installer - has a new package management system (Image Packaging System) - install most add-ons via internet, and can be updated via internet and the new package management system - is based on the opensolaris build 86 kernel (build 86 has known bugs when you try to install it in a PV domU on a xen-3.2 hypervisor) - Live-CD can be redistributed Solaris Express Community Editon (Solaris Express CE) http://www.opensolaris.org/os/downloads/ - Uses the old Solaris installer, comes on a DVD - Using SVr4 pkg utilities, everything needed for installation is on the DVD - Is released bi-weekly - currently using opensolaris build 90 kernel (so it should install in a PV domU on a xen-3.2 hypervisor without (?) issues) - Cannot be redistributed, so you must download it from sun.com.
I believe we got it:- bash-3.2# xm info host : Solaris87 release : 5.11 version : snv_91 machine : i86pc nr_cpus : 2 nr_nodes : 1 sockets_per_node : 1 cores_per_socket : 2 threads_per_core : 1 cpu_mhz : 3005 hw_caps : bfebfbff:20100800:00000000:00000140:0008e3fd:00000000:00000001 total_memory : 4095 free_memory : 128 xen_major : 3 xen_minor : 1 xen_extra : .4-xvm xen_caps : xen-3.0-x86_64 xen-3.0-x86_32p hvm-3.0-x86_32 hvm-3.0-x86_32p hvm-3.0-x86_64 xen_scheduler : credit xen_pagesize : 4096 platform_params : virt_start=0xffff800000000000 xen_changeset : Tue Jun 10 16:35:53 2008 +0400 15876:99a2426b9826 cc_compiler : gcc version 3.4.3 (csl-sol210-3_4-20050802) cc_compile_by : root cc_compile_domain : cc_compile_date : Tue Jun 10 18:02:46 MSD 2008 xend_config_format : 4 This message posted from opensolaris.org
************************ Make sure it works:- ************************ bash-3.2# xm list Name ID Mem VCPUs State Time(s) Domain-0 0 3886 2 r----- 64.5 Solaris10 1024 1 4373.8 Solaris87PVM 1024 1 306.7 bash-3.2# xm start Solaris87PVM bash-3.2# xm console Solaris87PVM v3.1.4-xvm chgset ''Tue Jun 10 16:35:53 2008 +0400 15876:99a2426b9826'' SunOS Release 5.11 Version snv_87 64-bit Copyright 1983-2008 Sun Microsystems, Inc. All rights reserved. Use is subject to license terms. Hostname: Solaris87pvm /dev/rdsk/c0d0s7 is clean Reading ZFS config: done. Solaris87pvm console login: root Password: Jun 10 18:32:35 Solaris87pvm login: ROOT LOGIN /dev/console Last login: Tue Jun 10 09:44:49 on console Sun Microsystems Inc. SunOS 5.11 snv_87 January 2008 -bash-3.2# This message posted from opensolaris.org
Thank you very much for your support and excellent advices. SunStudio 12 Download was already running , when you pointed me to install it from SXDE DVD. :-) This message posted from opensolaris.org
On Tue, Jun 10, 2008 at 10:16:44AM -0700, J??rgen Keil wrote:> > ---- building libvirt > ... > > checking whether /usr/sfw/bin/gmake sets $(MAKE)... yes > > checking for gcc... /opt/SUNWspro/SS11/bin/cc > > checking for C compiler default output file name... > > configure: error: C compiler cannot create executables > > I do remember... Not only is Sun Pro C required, you > also must have that SS11 directory inside SUNWspro.6707814 ./sunos.hg/bin/build-libvirt needs to work with the compilers downloaded from opensolaris.org regards john
El martes 10 de junio del 2008 a las 20:17:33, Juergen Keil escribió:> > Mmmmm... I am losing a little in this thread :-S What are the > > differences between 200805 and SNV*?> OpenSolaris 2008.05 (http://www.opensolaris.com/) > > - uses a Live-CD, from there you can start the installer > - has a new package management system (Image Packaging System) > - install most add-ons via internet, and can be updated via internet > and the new package management system > - is based on the opensolaris build 86 kernel > (build 86 has known bugs when you try to install it in a PV domU on > a xen-3.2 hypervisor) > - Live-CD can be redistributed> Solaris Express Community Editon (Solaris Express CE) > http://www.opensolaris.org/os/downloads/> - Uses the old Solaris installer, comes on a DVD > - Using SVr4 pkg utilities, everything needed for installation is on the DVD > - Is released bi-weekly > - currently using opensolaris build 90 kernel > (so it should install in a PV domU on a xen-3.2 hypervisor without (?) > issues) > - Cannot be redistributed, so you must download it from sun.com.Thanks for indicate the differences, Juergen. Trying download the single image DVD, I get the same problem mentioned before. The ISO image is empty. Somebody has the same problem? But I don''t have problems when trying to download the images of segment 1 and 2. They are necessary both? Regards, Daniel -- Daniel Bareiro - System Administrator Fingerprint: BFB3 08D6 B4D1 31B2 72B9 29CE 6696 BF1B 14E6 1D37 Powered by Debian GNU/Linux Lenny - Linux user #188.598
# cat *-a *-b > snv90.iso This message posted from opensolaris.org
SNV90 HVM at Xen 3.2.1 F8 Dom0 (both 64-bit) bfu''ed to SNV91. For now :- bash-3.2# uname -a SunOS dhcppc8 5.11 snv_91 i86pc i386 i86pc bash-3.2# bin/fork -E -C 200 -L -S -W -N fork_100 -B 100 -C 100 > /tmp/fork.output Running: fork_100 for 34.55362 seconds This message posted from opensolaris.org
El miércoles 11 de junio del 2008 a las 00:34:37, Boris Derzhavets escribió:> # cat *-a *-b > snv90.isoGood. I already have the single image from the concatenation of both archives. Then I tried to make the installation. sirius:/etc/xen# cat febo.cfg name = ''solaris'' memory = ''256'' disk = [ ''file:/mnt/snv90.iso,6:cdrom,r'', ''phy:vm1/febo,0,w'' ] vif = [ '''' ] bootloader = ''/usr/lib/xen-3.2-1/bin/pygrub'' kernel = ''/boot/platform/i86xpv/kernel/unix'' ramdisk = ''/boot/amd64/x86.miniroot'' extra = ''/platform/i86xpv/kernel/unix - nowin -B install_media=cdrom'' sirius:/etc/xen# xm create -c febo.cfg Using config file "./febo.cfg". sirius:/etc/xen# Error: (1, ''Internal error'', ''xc_dom_boot_domU_map: failed to mmap domU pages 0x826+0x176a6 [xenctl, rc=-1]\n'') Do you know what can originate this failure? The installation seems to stop here. Regards, Daniel -- Daniel Bareiro - System Administrator Fingerprint: BFB3 08D6 B4D1 31B2 72B9 29CE 6696 BF1B 14E6 1D37 Powered by Debian GNU/Linux Lenny - Linux user #188.598
I guess profile is incorrect:- sirius:/etc/xen# cat febo.cfg name = ''solaris'' # memory = ''256'' memory = 1024 # disk = [ ''file:/mnt/snv90.iso,6:cdrom,r'', ''phy:vm1/febo,0,w'' ] disk = [ ''file:/mnt/snv90.iso,6:cdrom,r'', ''phy:/vm1/febo,0,w'' ] vif = [ '' '' ] bootloader = ''/usr/lib/xen-3.2-1/bin/pygrub'' You picked up 32-bit kernel and 64-bit ramdisk Correct options bellow:- 1.) 64-bit kernel = ''/boot/platform/i86xpv/kernel/amd64/unix'' ramdisk = ''/boot/amd64/x86.miniroot'' 2.) 32-bit kernel = ''/boot/platform/i86xpv/kernel/unix'' ramdisk = ''/boot/x86.miniroot'' It''s for 32-bit kernel extra = ''/platform/i86xpv/kernel/unix - nowin -B install_media=cdrom'' Please, submit:- # xm info # brctl show outputs --- On Wed, 6/11/08, Daniel Bareiro <daniel-listas@gmx.net> wrote: From: Daniel Bareiro <daniel-listas@gmx.net> Subject: Re: [xen-discuss] OpenSolaris 2008.05 domU on Debian dom0 To: xen-discuss@opensolaris.org Date: Wednesday, June 11, 2008, 3:14 PM El miércoles 11 de junio del 2008 a las 00:34:37, Boris Derzhavets escribió: > # cat *-a *-b > snv90.iso Good. I already have the single image from the concatenation of both archives. Then I tried to make the installation. sirius:/etc/xen# cat febo.cfg name = ''solaris'' memory = ''256'' disk = [ ''file:/mnt/snv90.iso,6:cdrom,r'', ''phy:vm1/febo,0,w'' ] vif = [ '''' ] bootloader = ''/usr/lib/xen-3.2-1/bin/pygrub'' kernel = ''/boot/platform/i86xpv/kernel/unix'' ramdisk = ''/boot/amd64/x86.miniroot'' extra = ''/platform/i86xpv/kernel/unix - nowin -B install_media=cdrom'' sirius:/etc/xen# xm create -c febo.cfg Using config file "./febo.cfg". sirius:/etc/xen# Error: (1, ''Internal error'', ''xc_dom_boot_domU_map: failed to mmap domU pages 0x826+0x176a6 [xenctl, rc=-1]\n'') Do you know what can originate this failure? The installation seems to stop here. Regards, Daniel -- Daniel Bareiro - System Administrator Fingerprint: BFB3 08D6 B4D1 31B2 72B9 29CE 6696 BF1B 14E6 1D37 Powered by Debian GNU/Linux Lenny - Linux user #188.598_______________________________________________ xen-discuss mailing list xen-discuss@opensolaris.org
On Wednesday, 11 June 2008 12:42:46 -0700, Boris Derzhavets wrote:> I guess profile is incorrect:- > > sirius:/etc/xen# cat febo.cfg > name = ''solaris'' > > # memory = ''256'' > memory = 1024I don''t have 1 GiB available for this domU, for that reason I put 256 MIB :-)> You picked up 32-bit kernel and 64-bit ramdisk > Correct options bellow:- > 1.) 64-bit > kernel = ''/boot/platform/i86xpv/kernel/amd64/unix'' > > ramdisk = ''/boot/amd64/x86.miniroot'' > > 2.) 32-bit > kernel = ''/boot/platform/i86xpv/kernel/unix'' > ramdisk = ''/boot/x86.miniroot''Do''h! Small detail. You''re right :-/ The config file is thus: name = ''solaris'' memory = 256 disk = [ ''file:/mnt/snv90.iso,6:cdrom,r'', ''phy:vm1/febo,0,w'' ] vif = [ '''' ] bootloader = ''/usr/lib/xen-3.2-1/bin/pygrub'' kernel = ''/boot/platform/i86xpv/kernel/unix'' ramdisk = ''/boot/x86.miniroot'' extra = ''/platform/i86xpv/kernel/unix - nowin -B install_media=cdrom'' Then, I create the domU: sirius:/etc/xen# xm create -c febo.cfg Using config file "./febo.cfg". v3.2-1 chgset ''unavailable'' Started domain solaris SunOS Release 5.11 Version snv_90 32-bit Copyright 1983-2008 Sun Microsystems, Inc. All rights reserved. Use is subject to license terms. do_bop_phys_alloc(0xf5053940f5053930, 0xf4c0ef54f50539a8) Out of memory Press any key to reboot. [ENTER] Resetting... panic[cpu0]sirius:/etc/xen#> Please, submit:- > # xm info > # brctl show > outputssirius:/etc/xen# xm info host : sirius release : 2.6.18-6-xen-686 version : #1 SMP Sat May 24 14:16:03 UTC 2008 machine : i686 nr_cpus : 1 nr_nodes : 1 cores_per_socket : 1 threads_per_core : 1 cpu_mhz : 2200 hw_caps : 078bfbff:e3d3fbff:00000000:00000010:00000001:00000001:00000001 total_memory : 1271 free_memory : 426 node_to_cpu : node0:0 xen_major : 3 xen_minor : 2 xen_extra : -1 xen_caps : xen-3.0-x86_32p xen_scheduler : credit xen_pagesize : 4096 platform_params : virt_start=0xf5800000 xen_changeset : unavailable cc_compiler : gcc version 4.1.2 20061115 (prerelease) (Debian 4.1.1-21) cc_compile_by : fs cc_compile_domain : debian.org cc_compile_date : Mon Mar 10 15:50:27 UTC 2008 xend_config_format : 4 sirius:/etc/xen# brctl show bridge name bridge id STP enabled interfaces xenbr0 8000.feffffffffff no vif0.0 peth0 vif1.0 vif2.0 vif3.0 Thanks for your reply, Boris. Regards, Daniel -- Daniel Bareiro - System Administrator Fingerprint: BFB3 08D6 B4D1 31B2 72B9 29CE 6696 BF1B 14E6 1D37 Powered by Debian GNU/Linux Lenny - Linux user #188.598
> > 2.) 32-bit > > kernel = ''/boot/platform/i86xpv/kernel/unix'' > > ramdisk = ''/boot/x86.miniroot'' > > Do''h! Small detail. You''re right :-/ > > The config file is thus: > > name = ''solaris'' > memory = 256 > disk = [ ''file:/mnt/snv90.iso,6:cdrom,r'', > ''phy:vm1/febo,0,w'' ] > vif = [ '''' ] > bootloader = ''/usr/lib/xen-3.2-1/bin/pygrub'' > kernel = ''/boot/platform/i86xpv/kernel/unix'' > ramdisk = ''/boot/x86.miniroot'' > extra = ''/platform/i86xpv/kernel/unix - nowin -B > install_media=cdrom'' > > Then, I create the domU: > > sirius:/etc/xen# xm create -c febo.cfg > Using config file "./febo.cfg". > v3.2-1 chgset ''unavailable'' > Started domain solaris > SunOS Release 5.11 Version snv_90 32-bit > Copyright 1983-2008 Sun Microsystems, Inc. All rights reserved. > Use is subject to license terms. > do_bop_phys_alloc(0xf5053940f5053930, 0xf4c0ef54f50539a8) Out of memory > > Press any key to reboot. > [ENTER] > Resetting...Try to allocate more memory for the domain. Try it with 384 mb or 512 mb. After solaris is installed in the PV domU, you can try to reduce the memory to 256 mb, This message posted from opensolaris.org
On Thursday, 12 June 2008 03:26:01 -0700, Jürgen Keil wrote:> Try to allocate more memory for the domain. Try it with 384 mb or 512 > mb.> After solaris is installed in the PV domU, you can try to reduce the > memory to 256 mb,Good. Now using 384 MiB, the installer begins. It is a progress :-) But doesn''t seem to up the networking via DHCP: network interface name ''eth0'' replaced with ''xnf0'' --+ NOTICE: xdf@6: failed to read feature-barrier | This is normal message NOTICE: xdf@0: failed to read feature-barrier | in a virtual environment? --+ Attempting to configure interface xnf0... Skipped interface xnf0 [ Continuing with the installation... ] Could not contact a dhcp server on the network interface xnf0 These are the network lines in config file: vif = [ ''bridge=xenbr0'' ] dhcp = ''dhcp'' Thanks for your reply, Jürgen. Regards, Daniel -- Daniel Bareiro - System Administrator Fingerprint: BFB3 08D6 B4D1 31B2 72B9 29CE 6696 BF1B 14E6 1D37 Powered by Debian GNU/Linux Lenny - Linux user #188.598
On 12 Jun 2008, at 5:36pm, Daniel Bareiro wrote:> Good. Now using 384 MiB, the installer begins. It is a progress :-) > But > doesn''t seem to up the networking via DHCP: > > network interface name ''eth0'' replaced with ''xnf0''This is normal.> --+ > NOTICE: xdf@6: failed to read feature-barrier | This is normal message > NOTICE: xdf@0: failed to read feature-barrier | in a virtual > environment? > --+These are normal - your Debian dom0 disk backend doesn''t support the barrier feature.> Attempting to configure interface xnf0... > Skipped interface xnf0This means that RARP/bootparams failed (normal unless you have a Jumpstart environment in place).> [ Continuing with the installation... ] > > Could not contact a dhcp server on the network interface xnf0Obviously not good. Max is fixing a bug caused by a interrupt race condition when bringing up xnf0 which is one possible explanation for this.