Bug 840300
Summary: | 3.1 beta3 - [vdsm][bootstrap]: we do not fail installation when failing to create rhevm network | ||||||||
---|---|---|---|---|---|---|---|---|---|
Product: | Red Hat Enterprise Linux 6 | Reporter: | Dafna Ron <dron> | ||||||
Component: | vdsm | Assignee: | Douglas Schilling Landgraf <dougsland> | ||||||
Status: | CLOSED ERRATA | QA Contact: | GenadiC <gcheresh> | ||||||
Severity: | medium | Docs Contact: | |||||||
Priority: | medium | ||||||||
Version: | 6.3 | CC: | abaron, acathrow, alonbl, bazulay, cpelland, danken, dfediuck, dyasny, iheim, ilvovsky, jbiddle, lpeer, mavital, Rhev-m-bugs, sgrinber, ykaul | ||||||
Target Milestone: | beta | ||||||||
Target Release: | --- | ||||||||
Hardware: | x86_64 | ||||||||
OS: | Linux | ||||||||
Whiteboard: | infra network | ||||||||
Fixed In Version: | vdsm-4.9.6-38.0 | Doc Type: | Bug Fix | ||||||
Doc Text: |
Previously, if the libvirt daemon crashed during Red Hat Enterprise Virtualization host setup, the installation would continue regardless. This resulted in a failure to create a bridge to the Red Hat Enterprise Virtualization Manager, making the host unresponsive and unusable. Now, a condition has been set so if the libvirt daemon crashes, the installation fails and can be started again properly.
|
Story Points: | --- | ||||||
Clone Of: | Environment: | ||||||||
Last Closed: | 2012-12-04 19:02:42 UTC | Type: | Bug | ||||||
Regression: | --- | Mount Type: | --- | ||||||
Documentation: | --- | CRM: | |||||||
Verified Versions: | Category: | --- | |||||||
oVirt Team: | --- | RHEL 7.3 requirements from Atomic Host: | |||||||
Cloudforms Team: | --- | Target Upstream Version: | |||||||
Embargoed: | |||||||||
Attachments: |
|
This is a bootstrapping issue, moved to VDSM Patch upstream available: http://gerrit.ovirt.org/6875 Hello, I agree that current solution would to fail the bootstrap process. Later on we can discuss improvements to the process. As I discussed previously with dan, the use of libvirt during bootstrap is something that I personally dislike. It is caused by the need to use its persisted database to store the bridge. While libvirt is actually our slave and vdsm is the master, so any configuration should be persisted by vdsm and pushed to libvirt upon vdsm startup. The fact that libvirt has capabilities to store information, should not imply that it is a database for vdsm, the fact that it is implements persist data store, does not imply that our design should depend on this. If I would design a pure box diagram of components, I would probably put libvirt as a stateless components [although it may store state, but this state gets overridden each startup]. In the above case, all we need to do is to mark the existence of this bridge, then when vdsm will start it will configure the libvirt properly. So no failure during bootstrap. Alon. Created attachment 620284 [details]
vdsm bootstrap log
The installation finished with success (although the rhevm bridge was not created) Expected result: The installation should fail. The log is attached Patch upstream available for review: http://gerrit.ovirt.org/#/c/8345/ Downstream version: https://gerrit.eng.lab.tlv.redhat.com/#/c/2480/ Verified in SI24 (vdsm-4.9.6-41.0.el6_3.x86_64) In order to verify we changed the ifup to non executable. Since the problem described in this bug report should be resolved in a recent advisory, it has been closed with a resolution of ERRATA. For information on the advisory, and where to find the updated files, follow the link below. If the solution does not work for you, open a new bug report. http://rhn.redhat.com/errata/RHSA-2012-1508.html |
Created attachment 598311 [details] logs Description of problem: libvirt creashed right after it was started (probably because of dbus). addNetwork failed and we continues installation without having rhevm network. host becomes non-operational after reboot. Version-Release number of selected component (if applicable): How reproducible: Steps to Reproduce: 1. stop dbus before installing host 2. 3. Actual results: libvirt will crash after it starts. we will not fail host install and continue without configuring rhevm interface and host will become non-operational after reboot Expected results: if we fail to configure rhevm interface we should fail host install. Additional info: full logs attached Sun, 15 Jul 2012 15:41:16 DEBUG Starting libvirtd daemon: [ OK ] un, 15 Jul 2012 15:41:16 DEBUG libvir: RPC error : Failed to connect socket to '/var/run/libvirt/libvirt-sock': No such file or directory Traceback (most recent call last): File "/usr/lib64/python2.6/runpy.py", line 122, in _run_module_as_main "__main__", fname, loader, pkg_name) File "/usr/lib64/python2.6/runpy.py", line 34, in _run_code exec code in run_globals File "/usr/share/vdsm/configNetwork.py", line 1202, in <module> main() File "/usr/share/vdsm/configNetwork.py", line 1171, in main addNetwork(bridge, **kwargs) File "/usr/share/vdsm/configNetwork.py", line 582, in addNetwork _netinfo = NetInfo() File "/usr/lib64/python2.6/site-packages/vdsm/netinfo.py", line 307, in __init__ _netinfo = get() File "/usr/lib64/python2.6/site-packages/vdsm/netinfo.py", line 239, in get nets = networks() File "/usr/lib64/python2.6/site-packages/vdsm/netinfo.py", line 70, in networks conn = libvirtconnection.get() File "/usr/lib64/python2.6/site-packages/vdsm/libvirtconnection.py", line 113, in get