9:03 p.m.
Hi all,
I have done a simple checking for Xen and VMX on FC5-test2. Basically, I can create VMX on
FC5-test2 IA32. But FC5-test2 IA32e (x86-64) doesn't provide Xen and Xen0 kernel, so
didn't do Xen and VMX testing for it.
Beside of testing Xen and VMX, I also found some issues about installation, which I have
updated the bug #179078 as a NB. The following is the testing results only for Xen and
VMX.
=========================
1. *** IA32E hasn't RPMS about Xen hypervisor and Xen0/XenU kernels. __FAIL__
2. *** In order to boot up IA32 Xen0, I have to add pci=noacpi as the boot
parameters of Xen0 kernel, or system will hang when initial PCI. I don't know if it is
only related to my Machine. __FAIL__
3. After install IA32 Xen RPMs, I can see /boot/grub/grub.conf has been changed
with Xen kernels. __PASS__
4. *** /usr/lib/xen/boot/vmxloader isn't in the RPM package, so VMX Guest
couldn't be created. I suspected build system hasn't bcc, it is from DEV86
package. __FAIL__
5. After copy a useable vmxloader to the destination, IA32 VMX Guest can be
created successfully! __PASS__
6. VMX has good network connection and Xwindows started. __PASS__
7. XenU can be created and works well. __PASS__
8. Couldn't get Xen Changeset by "xm info" or "xm dmesg".
9. The new kernel names of Xen/Xen0/XenU seemed a little confused. E.g. the Xen0 name is
like vmlinuz-2.6.15-1.29_FC5hypervisor. I thought hypervisor is for Xen.gz. Xen0 is the
Service OS.
==========================
Any comments please let me know. But I will have 8 days holiday from tomorrow, so emails
reply might be slow and occasional.
[FYI] We are planning to do more testing for FC5-test3.
Best Regards,
Yongkang (Kangkang) 永康
10:03 a.m.
On Fri, 2006-01-27 at 11:03 +0800, You, Yongkang wrote:
1. *** IA32E hasn't RPMS about Xen hypervisor and Xen0/XenU
kernels. __FAIL__
Getting x86_64 building involved a couple of chicken sacrifices I think.
Stephen was making good progress yesterday, though, so x86_64 will
hopefully show up for test3.
2. *** In order to boot up IA32 Xen0, I have to add pci=noacpi as
the boot parameters of Xen0 kernel, or system will hang when
initial PCI. I don't know if it is only related to
my Machine. __FAIL__
Please file a bug including some actual details about your hardware
instead of just vague statements like this.
4. *** /usr/lib/xen/boot/vmxloader isn't in the RPM package,
so VMX Guest couldn't be created. I suspected build system hasn't
bcc, it is from DEV86 package. __FAIL__
Although it would be great if someone would actually port this stuff to
not use ancient, barely maintained toolchains and/or the approach of
actually distributing the firmware in the repository (as done by Bochs,
which is the upstream for this stuff), I went ahead and got dev86 to
build on x86_64 so that vmxloader is in later xen packages.
9. The new kernel names of Xen/Xen0/XenU seemed a little confused.
E.g. the Xen0 name is like vmlinuz-2.6.15-1.29_FC5hypervisor. I
thought hypervisor is for Xen.gz. Xen0 is the Service OS.
The kernel-xen-hypervisor package includes both the Xen hypervisor and
the dom0 kernel as these two are tightly intertwined.
Jeremy
7:10 a.m.
New subject: Network Problem after todays update
Hi ,
i did an update today from kernel-xen-2.6.18-1.2869.fc6 to
kernel-xen-2.6.19-1.2895.fc6 and from xen-3.0.3-1.fc6 to xen-3.0.3-3.fc6 .
the upgrade itself went fine, and all on dom_0 works as expected.
But i have now trouble with my guetsts Network. All guests are connecting
to the outside world over NAT.
i can ssh between all guests as well as between dom_0 and all guests.
i can ping each box outside the dom_0 (my dns server for insteance) but i
can't ssh or ftp to any box outside my internal dom_0 network, it always
timed out.
looking at a tcp dump it marks the frames i send from the guests as
Checksum Errors.
any ideas ? should i open a bugzilla entry for this ?
Sven
7:19 a.m.
New subject: Network Problem after todays update
sorry, there was a small typo in my mail, i upgraded from
kernel-xen-2.6.18-1.2868.fc6 to kernel-xen-2.6.19-1.2895.fc6 not from
-1.2869 ..
btw. going back to kernel .2868 fixes it, so there must be a bug added
between .2868 and .2895 i will test with .2869 later today ..
Sven
Sven Oehme/Germany/IBM@IBMDE
Sent by: fedora-xen-bounces(a)redhat.com
01/12/07 02:10 PM
To
fedora-xen(a)redhat.com
cc
Subject
[Fedora-xen] Network Problem after todays update
Hi ,
i did an update today from kernel-xen-2.6.18-1.2869.fc6 to
kernel-xen-2.6.19-1.2895.fc6 and from xen-3.0.3-1.fc6 to xen-3.0.3-3.fc6 .
the upgrade itself went fine, and all on dom_0 works as expected.
But i have now trouble with my guetsts Network. All guests are connecting
to the outside world over NAT.
i can ssh between all guests as well as between dom_0 and all guests.
i can ping each box outside the dom_0 (my dns server for insteance) but i
can't ssh or ftp to any box outside my internal dom_0 network, it always
timed out.
looking at a tcp dump it marks the frames i send from the guests as
Checksum Errors.
any ideas ? should i open a bugzilla entry for this ?
Sven--
Fedora-xen mailing list
Fedora-xen(a)redhat.com
https://www.redhat.com/mailman/listinfo/fedora-xen
9:15 a.m.
Hi,
On Fri, 2006-01-27 at 11:03 +0800, You, Yongkang wrote:
I have done a simple checking for Xen and VMX on FC5-test2.
Basically, I can create VMX on FC5-test2 IA32. But FC5-test2 IA32e (x86-64) doesn't
provide Xen and Xen0 kernel, so didn't do Xen and VMX testing for it.
Beside of testing Xen and VMX, I also found some issues about installation, which I have
updated the bug #179078 as a NB. The following is the testing results only for Xen and
VMX.
=========================
1. *** IA32E hasn't RPMS about Xen hypervisor and Xen0/XenU kernels. __FAIL__
It's coming. I hope. The upstream x86_64 stuff from linux-2.6-merge
doesn't actually compile at the moment (though that's easy to fix), and
once compiled it fails in different, compiler-dependent ways on some of
my hardware. I've got a minimally stable boot now on a different box,
but the upstream base in this case is proving hard to work with.
2. *** In order to boot up IA32 Xen0, I have to add pci=noacpi as the
boot
parameters of Xen0 kernel, or system will hang when initial PCI. I don't know if it
is only related to my Machine.
It's certainly hardware-dependent, but you're not the only person
reporting the need for such options. Please open a bugzilla so that we
can track this one properly, as we'd really like to automate such
options as much as possible for the main release.
One thing that would be worth knowing is whether the main, non-xen
rawhide SMP kernel boots on this box. The UP kernel disables a bunch of
advanced hardware enablement for compatibility purposes, but the SMP one
should match Xen as closely as possible.
8. Couldn't get Xen Changeset by "xm info" or "xm
dmesg".
Not sure how easy that would be to add, since we don't build it out of
Mercurial directly but rather out of a tarballed copy from the CVS build
system. I'm being careful to add the changeset info to the changelog in
the rpm, though.
Thanks,
Stephen
6316
days inactive
6666
days old
4 comments
4 participants
participants (4)
-
Jeremy Katz -
Stephen C. Tweedie -
Sven Oehme -
You, Yongkang