pcs : update of a vm resource with remote-node returns an abnormal error
Issue
Let's suppose a vm-gdump0 resource already exist in the cib.xml:
pcs resource show vm-gdump0
Resource: vm-gdump0 (class=ocf provider=heartbeat type=VirtualDomain)
Attributes: config=/var/lib/vm_images/gdump/gdump.xml force_stop=false hypervisor=qemu:///system
Meta Attrs: failure-timeout=60 migration-threshold=2 remote-connect-timeout=90 remote-node=gdump0.x
Utilization: cpu=2 hv_memory=2048
Operations: monitor interval=60 on-fail=restart timeout=90 (vm-gdump0-monitor-interval-60)
start interval=0s timeout=120 (vm-gdump0-start-interval-0s)
stop interval=0s timeout=90 (vm-gdump0-stop-interval-0s)
Just execute an update of the resource:
/usr/sbin/pcs resource update vm-gdump0 config=/var/lib/vm_images/gdump/gdump.xml force_stop=false hypervisor=qemu:///system [^] op monitor interval=60 on-fail=restart timeout=90 op start interval=0s timeout=120 op stop interval=0s timeout=90 meta failure-timeout=60 migration-threshold=2 remote-connect-timeout=90 remote-node=gdump0.x target-role=Stopped
Warning: this command is not sufficient for creating a guest node, use 'pcs cluster node add-guest'
Error: 'gdump0.bullx' already exists
This error is abnormal as the vm-gdump0 already exists with its attribute remote-node=gdump0.x, and the update commands is just to add target-role=Stopped (update of target-role is for example)
Environment
RHEL 7.4
Subscriber exclusive content
A Red Hat subscription provides unlimited access to our knowledgebase, tools, and much more.