Hi all,
I just did a fresh install of FC7 to see if xen would work out of the box. Dom0 boots (with errors in the log - see below) and some Centos 4.4 domU boots but are unseable (after recompiling the initrd image with the xen driver preloaded) up to the famous long lasting error message "4gb seg fixup". At that point it is dead
My question: what id the status of Xen for Fc7? Will it ever fly? Is there a Linux distro that works out of the box ?
The last usable version I saw was for FC5 as of May 2006, after that all version were badly broken and never fixed, I tested ALL of them for fc5/fc6 and not a single one worked as published. Only home brewed compiled version would work after weeks of trial and error.
Here what the log says from my out of the box install:
[root@paine xen]# cat xend.log [2007-06-29 08:08:34 2567] INFO (SrvDaemon:331) Xend Daemon started [2007-06-29 08:08:34 2567] INFO (SrvDaemon:335) Xend changeset: unavailable. [2007-06-29 08:08:34 2567] INFO (SrvDaemon:342) Xend version: Unknown. [2007-06-29 08:08:35 2567] DEBUG (XendDomainInfo:126) XendDomainInfo.recreate({'max_vcpu_id': 0, 'cpu_time': 35199342574L, 'ssidref': 0, 'hvm': 0, 'shutdown_reason': 0, 'dying': 0, 'online_vcpus': 1, 'domid': 0, 'paused': 0, 'crashed': 0, 'running': 1, 'maxmem_kb': 4294967292L, 'shutdown': 0, 'mem_kb': 962164L, 'handle': [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0], 'blocked': 0, 'name': 'Domain-0'}) [2007-06-29 08:08:35 2567] INFO (XendDomainInfo:143) Recreating domain 0, UUID 00000000-0000-0000-0000-000000000000. at /local/domain/0 [2007-06-29 08:08:35 2567] DEBUG (XendDomainInfo:2022) Storing VM details: {'on_xend_stop': 'ignore', 'shadow_memory': '0', 'uuid': '00000000-0000-0000-0000-000000000000', 'on_reboot': 'restart', 'image': '(linux (kernel ))', 'on_poweroff': 'destroy', 'on_xend_start': 'ignore', 'on_crash': 'restart', 'xend/restart_count': '0', 'vcpus': '1', 'vcpu_avail': '1', 'name': 'Domain-0'} [2007-06-29 08:08:35 2567] DEBUG (XendDomainInfo:873) Storing domain details: {'name': 'Domain-0', 'console/limit': '1048576', 'vm': '/vm/00000000-0000-0000-0000-000000000000', 'domid': '0', 'cpu/0/availability': 'online', 'memory/target': '962164', 'control/platform-feature-multiprocessor-suspend': '1'} [2007-06-29 08:08:35 2567] DEBUG (XendDomain:443) Adding Domain: 0 [2007-06-29 08:08:35 2567] DEBUG (XendDomain:379) number of vcpus to use is 0 [2007-06-29 08:08:35 2567] INFO (SrvServer:180) unix path=/var/lib/xend/xend-socket [2007-06-29 08:08:35 2567] DEBUG (XendDomainInfo:873) Storing domain details: {'name': 'Domain-0', 'console/limit': '1048576', 'vm': '/vm/00000000-0000-0000-0000-000000000000', 'domid': '0', 'cpu/0/availability': 'online', 'memory/target': '962164', 'control/platform-feature-multiprocessor-suspend': '1'} [2007-06-29 08:08:35 2567] DEBUG (XendDomainInfo:957) XendDomainInfo.handleShutdownWatch [2007-06-29 08:08:35 2567] WARNING (XendAPI:672) API call: VBD.set_device not found [2007-06-29 08:08:35 2567] WARNING (XendAPI:672) API call: VBD.set_mode not found [2007-06-29 08:08:35 2567] WARNING (XendAPI:672) API call: VBD.set_type not found [2007-06-29 08:08:35 2567] WARNING (XendAPI:672) API call: VM.get_auto_power_on not found [2007-06-29 08:08:35 2567] WARNING (XendAPI:672) API call: VM.set_auto_power_on not found [2007-06-29 08:08:35 2567] WARNING (XendAPI:672) API call: VM.set_VCPUs_max not found [2007-06-29 08:08:35 2567] WARNING (XendAPI:672) API call: VM.set_VCPUs_at_startup not found [2007-06-29 08:08:35 2567] WARNING (XendAPI:672) API call: debug.get_all not found [2007-06-29 08:08:35 2567] WARNING (XendAPI:672) API call: console.get_other_config not found [2007-06-29 08:08:35 2567] WARNING (XendAPI:672) API call: console.set_other_config not found [2007-06-29 08:08:35 2567] WARNING (XendAPI:672) API call: VIF.get_network not found [2007-06-29 08:08:35 2567] WARNING (XendAPI:672) API call: VIF.set_device not found [2007-06-29 08:08:35 2567] WARNING (XendAPI:672) API call: VIF.set_MAC not found [2007-06-29 08:08:35 2567] WARNING (XendAPI:672) API call: VIF.set_MTU not found [2007-06-29 08:08:35 2567] WARNING (XendAPI:672) API call: session.get_all_records not found [2007-06-29 08:08:35 2567] WARNING (XendAPI:672) API call: event.get_record not found [2007-06-29 08:08:35 2567] WARNING (XendAPI:672) API call: event.get_all not found [2007-06-29 08:08:46 2567] INFO (XMLRPCServer:149) Opening Unix domain socket XML-RPC server on /var/run/xend/xen-api.sock; authentication has been disabled for this server. [2007-06-29 08:08:46 2567] INFO (XMLRPCServer:149) Opening Unix domain socket XML-RPC server on /var/run/xend/xmlrpc.sock.
Regards,
Frank Morane wrote:
Hi all,
I just did a fresh install of FC7 to see if xen would work out of the box. Dom0 boots (with errors in the log - see below) and some Centos 4.4 domU boots but are unseable (after recompiling the initrd image with the xen driver preloaded) up to the famous long lasting error message "4gb seg fixup". At that point it is dead
I don't understand what you mean by "dead". Does the dom0 hang?
My question: what id the status of Xen for Fc7? Will it ever fly? Is there a Linux distro that works out of the box ?
I've been running FC7's Xen here quite well with a Centos 5 guest (amongst others). I haven't tried RHEL4 or Centos 4 as a guest though.
Rich.
Richard W.M. Jones wrote:
Frank Morane wrote:
Hi all,
I just did a fresh install of FC7 to see if xen would work out of the box. Dom0 boots (with errors in the log - see below) and some Centos 4.4 domU boots but are unseable (after recompiling the initrd image with the xen driver preloaded) up to the famous long lasting error message "4gb seg fixup". At that point it is dead
I don't understand what you mean by "dead". Does the dom0 hang?
No, just the domU. It does nothing more once the "4gb seg fixup" start spiting.
My question: what id the status of Xen for Fc7? Will it ever fly? Is there a Linux distro that works out of the box ?
I've been running FC7's Xen here quite well with a Centos 5 guest (amongst others). I haven't tried RHEL4 or Centos 4 as a guest though.
That is my next step to test BUT I have a bunch of older Centos to migrate so I would like to see them work on newer Xen version.
Thanks.
Rich.
Frank Morane wrote:
Richard W.M. Jones wrote:
Frank Morane wrote:
Hi all,
I just did a fresh install of FC7 to see if xen would work out of the box. Dom0 boots (with errors in the log - see below) and some Centos 4.4 domU boots but are unseable (after recompiling the initrd image with the xen driver preloaded) up to the famous long lasting error message "4gb seg fixup". At that point it is dead
I don't understand what you mean by "dead". Does the dom0 hang?
No, just the domU. It does nothing more once the "4gb seg fixup" start spiting.
If you have a directory /lib/tls in the guest, try moving it out of the way before the guest boots.
You could also try:
http://www.redhat.com/archives/fedora-xen/2006-November/msg00142.html
See also section 2.5.3 here:
http://www.cl.cam.ac.uk/research/srg/netos/xen/readmes/user/user.html#tex2ht...
Let us know if any of this works.
Rich.
Frank Morane wrote:
Hi all,
I just did a fresh install of FC7 to see if xen would work out of the box. Dom0 boots (with errors in the log - see below) and some Centos 4.4 domU boots but are unseable (after recompiling the initrd image with the xen driver preloaded) up to the famous long lasting error message "4gb seg fixup". At that point it is dead
My question: what id the status of Xen for Fc7? Will it ever fly? Is there a Linux distro that works out of the box ?
Hmm. I'm running a F7 dom0 with the following domUs: NetBSD 3.1, FC6, F7, and a very shaky Solaris 10 b44.
Xen 3.1 + F7 seems rock solid.
The last usable version I saw was for FC5 as of May 2006, after that all version were badly broken and never fixed, I tested ALL of them for fc5/fc6 and not a single one worked as published. Only home brewed compiled version would work after weeks of trial and error.
Here what the log says from my out of the box install:
Are you sure you're not missing the domU's console? I noticed that with at least the FC6 and F7 domUs, the boot output starts on the serial console, then transfers to the graphical console. I'm not familiar with the Centos domU's behavior, but it might be something to check.
Mike wrote:
Frank Morane wrote:
Hi all,
I just did a fresh install of FC7 to see if xen would work out of the box. Dom0 boots (with errors in the log - see below) and some Centos 4.4 domU boots but are unseable (after recompiling the initrd image with the xen driver preloaded) up to the famous long lasting error message "4gb seg fixup". At that point it is dead
My question: what id the status of Xen for Fc7? Will it ever fly? Is there a Linux distro that works out of the box ?
Hmm. I'm running a F7 dom0 with the following domUs: NetBSD 3.1, FC6, F7, and a very shaky Solaris 10 b44.
Xen 3.1 + F7 seems rock solid.
The last usable version I saw was for FC5 as of May 2006, after that all version were badly broken and never fixed, I tested ALL of them for fc5/fc6 and not a single one worked as published. Only home brewed compiled version would work after weeks of trial and error.
Here what the log says from my out of the box install:
Are you sure you're not missing the domU's console? I noticed that with at least the FC6 and F7 domUs, the boot output starts on the serial console, then transfers to the graphical console. I'm not familiar with the Centos domU's behavior, but it might be something to check.
I think this is a big problem, it happened to me.* Unfortunately, I was trying on FC6 Test 2/3 so I was told to get the F7 release version and try that.
There is no clear step by step how to on how to build GUI-less domUs for FC6 or F7 (or more frustratingly, no notices that it is not possible), so many people seem to be trying it and asking variations of the same question here, the IRC channels for Fedora and Xen, the libvir mailing list, the Xen mailing list, etc.
Can someone with the access rights put a clear unequivocal checklists or statements on the libvir site and the F7 site with the steps that work and/or don't work for creating domUs for normal people? Lots of people are in GUI-less situations and trying F7/Xen and failing unnecessarily, its bad PR for Xen and F7.
And kudos to those running everything under the sun on an F7 dom0, but can someone give those of us who aren't in the secret club a break and publish a full step-by-step of how you did it and some config files/settings. Did you use a copy of the DVD** exactly as it was or change some magical files after copying to disk? Has GUI-less domU installs or installs using a manually assigned IPv4 address ever worked for anyone using an F7 dom0, if so how?
* [Along with the problem that the installer was crashing when manually configuring an IPv4 address for an interface.] The released one still seems to want the subnet mask in 32s rather than 255s.
** I also didn't know that a remote repo would fail during domU installs until someone mentioned it among some other attempts at troubleshooting my install failures.
Thanks to anyone who can help with this.
On Fri, 29 Jun 2007 19:17:54 -0700 FCXen wrote: F> > Are you sure you're not missing the domU's console? I noticed that with F> > at least the FC6 and F7 domUs, the boot output starts on the serial F> > console, then transfers to the graphical console. I'm not familiar with F> > the Centos domU's behavior, but it might be something to check. F> F> I think this is a big problem, it happened to me.* Unfortunately, I was trying F> on FC6 Test 2/3 so I was told to get the F7 release version and try that. F> F> There is no clear step by step how to on how to build GUI-less domUs for FC6 F> or F7
I simply edit the xen config file and comment out the virtual frame buffer line (vfb = [...]), and it works fine w/out a GUI...
And kudos to those running everything under the sun on an F7 dom0, but can someone give those of us who aren't in the secret club a break and publish a full step-by-step of how you did it and some config files/settings.
Wrapper for virt-install worked on F7 dom0 for FC6 and F7 domU http://bewley.net/linux/fedora/xen/mkxen.sh
Did you use a copy of the DVD** exactly as it was or change some magical files after
I created a mirror by rsyncing another site. DVD copy should work fine.
copying to disk? Has GUI-less domU installs or installs using a manually assigned IPv4 address ever worked for anyone using an F7 dom0, if so how?
I used DHCP.
** I also didn't know that a remote repo would fail during domU installs until someone mentioned it among some other attempts at troubleshooting my install failures.
My FC6 install was fine against my personal repo over HTTP. My F7 was fine, but quite quite slow. I could watch the http access log and tell it was still working though.
My F7 kickstart file which hasn't been optimized but basically it looks like this:
## upgrade or install? install # upgrade ## comment out if you want the machine to wait for you to reboot manually reboot ## install source on URL or CDROM? url --url http://ks/fedora/linux/releases/7/Fedora/x86_64/os #cdrom ## use text mode install since i'll be spying on the serial console text skipx lang en_US.UTF-8 keyboard us timezone America/Los_Angeles ## network setup network --device eth0 --bootproto dhcp #network --device eth1 --onboot no --bootproto dhcp ## we'll redo firewalling by hand later firewall --enabled --port=22:tcp ## this could be annoying for now so leave it off selinux --disabled authconfig --enableshadow --enablemd5 ## change this after install rootpw something ################################################################################ # Setup the disk drives. # Clear the Master Boot Record zerombr # nuke all existing partitions clearpart --all --initlabel # raw partition for boot and a LVM physical volume on remainder part /boot --fstype ext3 --size=100 part pv.01 --size=1 --grow # LVM setup on physical volume volgroup VG0 pv.01 logvol / --fstype=ext3 --name=LVRoot --vgname=VG0 --size=512 logvol /var --fstype=ext3 --name=LVVar --vgname=VG0 --size=1024 logvol /var/www --fstype=ext3 --name=LVVarWww --vgname=VG0 --size=2048 logvol /usr --fstype=ext3 --name=LVUsr --vgname=VG0 --size=2048 logvol /home --fstype=ext3 --name=LVHome --vgname=VG0 --size=100 logvol swap --fstype swap --name=LVSwap --vgname=VG0 --size=512 # install grub bootloader --location=mbr --driveorder=xvda --append="console=xvc0" %packages vim-enhanced ################################################################################ # Final Configuration %post --nochroot cp /tmp/ks.cfg /mnt/sysimage/root/install-ks.cfg cp /proc/cmdline /mnt/sysimage/root/install-cmdline
%post echo xvc0 >> /etc/securetty # this was done auto on FC6 domU echo "co:2345:respawn:/sbin/agetty xvc0 9600 vt100-nav" >> /etc/inittab # we'll use these values in the extended final config EMAIL=root@mail DIST=f7 MAC_ADDR=`ifconfig eth0 | grep HWaddr | \ sed -e 's/^.*HWaddr ([A-Fa-f0-9:]*).*$/\1/; s/:/-/g;'` # put /tmp on swap cuz it's fast and junk goes away on reboot echo -e "none\t\t\t/tmp\t\t\ttmpfs\tdefaults\t0 0" >> /etc/fstab # setup root's profile echo 'alias vi=vim' >> /root/.bash_profile # I want to know about things... echo -e "root:\t\t$EMAIL" >> /etc/aliases newaliases # tell daddy we are all done cat /root/install.log /root/install-ks.cfg \ | mail -s "${DIST} ks install ${MAC_ADDR}" $EMAIL
-- Dale Bewley - Unix Administrator - Shields Library - UC Davis GPG: 0xB098A0F3 0D5A 9AEB 43F4 F84C 7EFD 1753 064D 2583 B098 A0F3
Dale Bewley wrote:
And kudos to those running everything under the sun on an F7 dom0, but can someone give those of us who aren't in the secret club a break and publish a full step-by-step of how you did it and some config files/settings.
Wrapper for virt-install worked on F7 dom0 for FC6 and F7 domU http://bewley.net/linux/fedora/xen/mkxen.sh
Wow, awesome, thanks. I will try this soon. Not interactive and not static IP but great for server setups.
Mike wrote:
Frank Morane wrote:
Hi all,
I just did a fresh install of FC7 to see if xen would work out of the box. Dom0 boots (with errors in the log - see below) and some Centos 4.4 domU boots but are unseable (after recompiling the initrd image with the xen driver preloaded) up to the famous long lasting error message "4gb seg fixup". At that point it is dead
My question: what id the status of Xen for Fc7? Will it ever fly? Is there a Linux distro that works out of the box ?
Hmm. I'm running a F7 dom0 with the following domUs: NetBSD 3.1, FC6, F7, and a very shaky Solaris 10 b44.
Xen 3.1 + F7 seems rock solid.
The last usable version I saw was for FC5 as of May 2006, after that all version were badly broken and never fixed, I tested ALL of them for fc5/fc6 and not a single one worked as published. Only home brewed compiled version would work after weeks of trial and error.
Here what the log says from my out of the box install:
Are you sure you're not missing the domU's console? I noticed that with at least the FC6 and F7 domUs, the boot output starts on the serial console, then transfers to the graphical console. I'm not familiar with the Centos domU's behavior, but it might be something to check.
Ah!? This could be it, I will dig to unearth this topic as if what you say is right, this could be it.
Thanks
Mike wrote:
Frank Morane wrote:
Hi all,
I just did a fresh install of FC7 to see if xen would work out of the box. Dom0 boots (with errors in the log - see below) and some Centos 4.4 domU boots but are unseable (after recompiling the initrd image with the xen driver preloaded) up to the famous long lasting error message "4gb seg fixup". At that point it is dead
My question: what id the status of Xen for Fc7? Will it ever fly? Is there a Linux distro that works out of the box ?
Hmm. I'm running a F7 dom0 with the following domUs: NetBSD 3.1, FC6, F7, and a very shaky Solaris 10 b44.
Xen 3.1 + F7 seems rock solid.
The last usable version I saw was for FC5 as of May 2006, after that all version were badly broken and never fixed, I tested ALL of them for fc5/fc6 and not a single one worked as published. Only home brewed compiled version would work after weeks of trial and error.
Here what the log says from my out of the box install:
Are you sure you're not missing the domU's console? I noticed that with at least the FC6 and F7 domUs, the boot output starts on the serial console, then transfers to the graphical console. I'm not familiar with the Centos domU's behavior, but it might be something to check.
Hi,
Ok I did add the console as explained in another post BUT it did not solve the issue:
I added this line in /etc/inittab
# Required by Xen co:2345:respawn:/sbin/mingetty console
I still can't get a console BUT I can access via ssh the booted domU; then I saw nash using 95% of the cpu and I killed it.
The problem is that this is far from what I want since I cannot access the domU via a console, when I try I get these:
printk: 185 messages suppressed. 4gb seg fixup, process ntpd (pid 1230), cs:ip 73:008ac17a printk: 17121 messages suppressed. 4gb seg fixup, process ntpd (pid 1230), cs:ip 73:008ac17a printk: 183 messages suppressed. 4gb seg fixup, process ntpd (pid 1230), cs:ip 73:008ac17a
and yes I have the "hwcap 0 nosegneg" in /etc/ld.so.conf. There is critical information missing to get older generation (Centos 4) domU booting on late FC5, FC6 and FC7.
Anyway, any additional hints will be appreciated.
Regards,
Frank Morane wrote:
I still can't get a console BUT I can access via ssh the booted domU; then I saw nash using 95% of the cpu and I killed it.
The problem is that this is far from what I want since I cannot access the domU via a console, when I try I get these:
printk: 185 messages suppressed. 4gb seg fixup, process ntpd (pid 1230), cs:ip 73:008ac17a
[...]
and yes I have the "hwcap 0 nosegneg" in /etc/ld.so.conf. There is critical information missing to get older generation (Centos 4) domU booting on late FC5, FC6 and FC7.
nosegneg only works if the kernel supports it, and as far as I can find out, this capability was only added relatively recently (earlier this year).
http://www.uwsg.iu.edu/hypermail/linux/kernel/0704.2/2872.html
In addition, this flags affects how (or rather, which) glibc is dynamically linked in to the binary. /sbin/nash is statically linked, but /usr/sbin/ntpd is dynamically linked (at least on FC6, and I expect on FC5 too).
So if you're getting the above messages concerning ntpd, then it sounds a lot like you're still linking to a version of glibc which is using negative offsets with segment register overrides, which is something that Xen has to emulate, slowly.
As to how to fix this: In Debian the fix is well-known. Rename the directory /lib/tls (on 64 bit Fedora: /lib64/tls) as /lib/tls.disabled, and reboot the domU.
Rich.
Also you might want to read through and add something to these bugs:
https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=243756 https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=244730
Rich.
xen@lists.stg.fedoraproject.org