[Users] Big issue with oVirt
Dear, We have updated our oVirt installation to 3.3.1 and now we have the following problem: When we boot a VM: He doesn't automatically chooses a hypervisor, he always takes the same But a larger issue is this one: 2013-11-27 12:37:14,528 WARN [org.ovirt.engine.core.vdsbroker.VdsManager] (DefaultQuartzScheduler_Worker-1) Failed to refresh VDS , vds = 2a8b1829-2235-44db-be12-ac71dc6777f8 : Vostok, error = java.lang.NullPointerException, continuing.: java.lang.NullPointerException Vostok is the server where we try to boot the VM, sometimes the VM comes up, but most of the time they don't... And if I would say to boot the VM on a specific other Hypervisor, he will ga ve the same error, only with the other host(hypervisor) in the error log. Any Idea's? Kind regards, Koen
On Wed, Nov 27, 2013 at 3:23 PM, Itamar Heim wrote:
did you VMs have the baloon device enabled?
I reported this: As for the current cluster config: - optimization : none (disable memory page sharing) memory balloon optimization is not checked - resilience policy : migrate virtual machines - cluster policy : none Can I safely install ksm on f19 and then enable and test?
Hi, this was indeed a bug in our code. It happened when ballooning was enabled (Cluster / Optimization), but no balloon device was configured (Vm / Resource Allocation). This is fixed by http://gerrit.ovirt.org/#/c/21783/ and tracked by https://bugzilla.redhat.com/show_bug.cgi?id=1035297. We will merge it to master momentarily and then update the released version as well. Thanks for the report. -- Martin Sivák msivak@redhat.com Red Hat Czech RHEV-M SLA / Brno, CZ ----- Original Message -----
It could be another aspect related with a previous post of mine. See: http://lists.ovirt.org/pipermail/users/2013-November/018140.html In that thread we remained that Gilad Chaplik would have done some tests and come back.... Gianluca
Hello, We are in the process of testing oVirt and replacing citrix all together. A few things we are curious about but can't find any concrete information on: Can oVirt live migrate data (VM's, etc.) to another machine in the pool/cluster in the event of a hardware failure, if the data is stored locally on the machine? For this to be possible, does it have to be on a shared storage? Thank you, Neil
Il 27/11/2013 14:59, Neil Schulz ha scritto:
If you're going to use shared storage I think you may be interested in hosted engine feature. Take a look at: http://www.ovirt.org/Features/Self_Hosted_Engine and http://www.ovirt.org/Hosted_Engine_Howto (actually you'll need nightly repository for testing it)
-- Sandro Bonazzola Better technology. Faster innovation. Powered by community collaboration. See how it works at redhat.com
On 11/27/2013 03:59 PM, Neil Schulz wrote:
not yet. live storage migration is currently only supported between storage domains in same storage pool (data center) where the host see both domains. we are working to remove the limitations in the pool design, which will allow this in a future version. oVirt supports managing VMs in host local storage, but the more powerful features are around clusters of hosts using shared storage. shared storage these days can benefit from multiple hosts local storage, via technologies like gluster, though at this point i think most run the storage nodes separatly from the virtualization hosts. NFS shared storage seems to be the easiest, iscsi, FC and posixfs are also supported. HTH, Itamar
participants (7)
-
Gianluca Cecchi -
Itamar Heim -
Jakub Bittner -
Koen Vanoppen -
Martin Sivak -
Neil Schulz -
Sandro Bonazzola