On Thu, Jul 2, 2020 at 4:38 PM Ruslanas Gžibovskis <ruslanas@lpic.lt> wrote:
it is, i have image build failing. i can modify yaml used to create image. can you remind me which files it would be?


Right, I see that the patch must not be working fine for centos and the package is being installed from delorean repos in the log.  I guess it needs an entry to cover the centos 8 case (i'm checking with opstools maintainer).

As workaround I'd propose you to use the package from:

https://trunk.rdoproject.org/centos8-ussuri/component/cloudops/current-tripleo/

or alternatively applying some local patch to tripleo-puppet-elements.
 
and your question, "how it can impact kvm":

in image most of the packages get deployed from deloren repos. I believe part is from centos repos and part of whole packages in overcloud-full.qcow2 are from deloren. so it might have bit different minor version, that might be incompactible... at least it have happend for me previously with train release so i used tested ci fully from the beginning...
I might be for sure wrong.

Delorean repos contain only OpenStack packages, things like nova, etc... not kvm or things included in CentOS repos. KVM will always installed which should be installed from "Advanced Virtualization" repository. May you check what versions of qemu-kvm and libvirt you got installed into the overcloud-full image?, it should match with the versions in:

http://mirror.centos.org/centos/8/virt/x86_64/advanced-virtualization/Packages/q/

like qemu-kvm-4.2.0-19.el8.x86_64.rpm and libvirt-6.0.0-17.el8.x86_64.rpm
 

On Thu, 2 Jul 2020, 17:18 Alfredo Moralejo Alonso, <amoralej@redhat.com> wrote:


On Thu, Jul 2, 2020 at 3:59 PM Ruslanas Gžibovskis <ruslanas@lpic.lt> wrote:
by the way in CentOS8, here is an error message I receive when searching around

[stack@rdo-u ~]$ dnf list --enablerepo="*" --disablerepo "c8-media-BaseOS,c8-media-AppStream" | grep osops-tools-monitoring-oschecks
Errors during downloading metadata for repository 'rdo-trunk-ussuri-tested':
  - Status code: 403 for https://trunk.rdoproject.org/centos8-ussuri/current-passed-ci/repodata/repomd.xml (IP: 3.87.151.16)
Error: Failed to download metadata for repo 'rdo-trunk-ussuri-tested': Cannot download repomd.xml: Cannot download repodata/repomd.xml: All mirrors were tried
[stack@rdo-u ~]$ 


Yep, rdo-trunk-ussuri-tested repo included in the release rpm is disabled by default and not longer usable (i'll send a patch to retire it), don't enable it.

Sorry, I'm not sure how adding osops-tools-monitoring-oschecks may lead to install CentOS8 maintained kvm. BTW, i think that package should not be required in CentOS8:



 
On Thu, 2 Jul 2020 at 15:56, Ruslanas Gžibovskis <ruslanas@lpic.lt> wrote:
Hi All,

I have one idea, why it might be the issue.

during image creation step, I have hadd missing packets:
pacemaker-remote osops-tools-monitoring-oschecks pacemaker pcs
PCS thing can be found in HA repo, so I enabled it, but  "osops-tools-monitoring-oschecks" ONLY in delorene for CentOS8...

I believe that is a case...
so it installed non CentOS8 maintained kvm or some dependent packages....

How can I get  osops-tools-monitoring-oschecks from centos repos? it is last seen in CentOS7 repos....

$ yum list --enablerepo=* --disablerepo "c7-media" | grep osops-tools-monitoring-oschecks -A2
osops-tools-monitoring-oschecks.noarch     0.0.1-0.20191202171903.bafe3f0.el7
                                                                  rdo-trunk-train-tested
ostree-debuginfo.x86_64                    2019.1-2.el7           base-debuginfo
(undercloud) [stack@ironic-poc ~]$ 

can I somehow not include that package in image creation? OR if it is essential, can I create a different repo for that one?




On Wed, 1 Jul 2020 at 14:19, Ruslanas Gžibovskis <ruslanas@lpic.lt> wrote:
Hi all!

Here we go, we are in the second part of this interesting troubleshooting!

1) I have LogTool setup.Thank you Arkady.

2) I have user OSP to create instance, and I have used virsh to create instance.
2.1) OSP  way is failing in either way, if it is volume-based or image-based, it is failing either way.. [1] and [2]
2.2) when I create it using CLI: [0] [3]

any ideas what can be wrong? What options I should choose?
I have one network/vlan for whole cloud. I am doing proof of concept of remote booting, so I do not have br-ex setup. and I do not have br-provider.

There is my compute[5] and controller[6] yaml files, Please help, how it should look like so it would have br-ex and br-int connected? as br-int now is in UNKNOWN state. And br-ex do not exist.
As I understand, in roles data yaml, when we have tag external it should create br-ex? or am I wrong?

[3] http://paste.openstack.org/show/HExQgBo4MDxItAEPNaRR/ 


On Tue, 30 Jun 2020 at 16:02, Arkady Shtempler <ashtempl@redhat.com> wrote:
Hi all!

I was able to analyze the attached log files and I hope that the results may help you understand what's going wrong with instance creation.
You can find Log_Tool's unique exported Error blocks here: http://paste.openstack.org/show/795356/

Some statistics and problematical messages:
##### Statistics - Number of Errors/Warnings per Standard OSP log since: 2020-06-30 12:30:00 #####
Total_Number_Of_Errors --> 9
/home/ashtempl/Ruslanas/controller/neutron/server.log --> 1
/home/ashtempl/Ruslanas/compute/stdouts/ovn_controller.log --> 1
/home/ashtempl/Ruslanas/compute/nova/nova-compute.log --> 7

nova-compute.log
default default] Error launching a defined domain with XML: <domain type='kvm'>
368-2020-06-30 12:30:10.815 7 ERROR nova.compute.manager [req-87bef18f-ad3d-4147-a1b3-196b5b64b688 7bdb8c3bf8004f98aae1b16d938ac09b 69134106b56941698e58c61...
70dc50f] Instance failed to spawn: libvirt.libvirtError: internal error: qemu unexpectedly closed the monitor: 2020-06-30T10:30:10.182675Z qemu-kvm: error: failed to set MSR 0...
he monitor: 2020-06-30T10:30:10.182675Z qemu-kvm: error: failed to set MSR 0x48e to 0xfff9fffe04006172
_msrs: Assertion `ret == cpu->kvm_msr_buf->nmsrs' failed.
 [instance: 128f372c-cb2e-47d9-b1bf-ce17270dc50f] Traceback (most recent call last):
375-2020-06-30 12:30:10.815 7 ERROR nova.compute.manager [instance: 128f372c-cb2e-47d9-b1bf-ce17270dc50f]   File "/usr/lib/python3.6/site-packages/nova/vir...

server.log 
5821c815-d213-498d-9394-fe25c6849918', 'status': 'failed', 'code': 422} returned with failed status

ovn_controller.log
272-2020-06-30T12:30:10.126079625+02:00 stderr F 2020-06-30T10:30:10Z|00247|patch|WARN|Bridge 'br-ex' not found for network 'datacentre'

Thanks!

Compute nodes are baremetal or virtualized?, I've seen similar bug reports when using nested virtualization in other OSes.
baremetal. Dell R630 if to be VERY precise. 

Thank you, I will try. I also modified a file, and it looked like it relaunched podman container once config was changed. Either way, if I understand Linux config correctly, the default value for user and group is root, if commented out:
#user = "root"
#group = "root"

also in some logs, I saw, that it detected, that it is not AMD CPU :) and it is really not AMD CPU.


Just for fun, it might be important, here is how my node info looks.
  ComputeS01Parameters:
    NovaReservedHostMemory: 16384
    KernelArgs: "crashkernel=no rhgb"
  ComputeS01ExtraConfig:
    nova::cpu_allocation_ratio: 4.0
    nova::compute::libvirt::rx_queue_size: 1024
    nova::compute::libvirt::tx_queue_size: 1024
    nova::compute::resume_guests_state_on_host_boot: true
_______________________________________________




--
Ruslanas Gžibovskis
+370 6030 7030


--
Ruslanas Gžibovskis
+370 6030 7030
_______________________________________________
users mailing list
users@lists.rdoproject.org
http://lists.rdoproject.org/mailman/listinfo/users

To unsubscribe: users-unsubscribe@lists.rdoproject.org