0

I have 3 nodes with CephRBD + OpenNebula + Corosync + Pacemaker + Galera + crmsh.

Software: OpenNebula 4.14.2, CentOS 7.1

All works, but when I instantiate VM, a get an error and VM in PENDING status:

No system datastore meets capacity and SCHED_DS_REQUIREMENTS: CLUSTER_ID = 100 & !(PUBLIC_CLOUD = YES)

I found this report: CephRBD is for Images.

VM TEMPLATE

CONTEXT=[NETWORK="YES",SSH_PUBLIC_KEY="$USER[SSH_PUBLIC_KEY]"]
CPU="1"
DISK=[IMAGE="CentOS",IMAGE_UNAME="oneadmin",READONLY="yes"]
DISK=[IMAGE="VM_Image",IMAGE_UNAME="oneadmin"]
FEATURES=[ACPI="yes"]
GRAPHICS=[LISTEN="0.0.0.0",TYPE="VNC"]
HYPERVISOR="kvm"
LOGO="images/logos/centos.png"
MEMORY="512"
NIC=[NETWORK="Management",NETWORK_UNAME="oneadmin"]
OS=[BOOT="cdrom"]
SCHED_REQUIREMENTS="CLUSTER_ID=\"100\""

I have another single host without clustering, and there are no this problems. OpenNebula 4.12.2

Where do I have a mistake?

1 Answer 1

0

Each cluster needs to have (at least) a system DS. That scheduler message means that either your cluster 100 does not have a system DS, or the system DS found does not have enough capacity for the pending VM.

If there is a system DS in that cluster, it may not be properly monitored, showing a storage capacity of 0.

3
  • So, I must have a shared system DS at all nodes too? Apr 11, 2016 at 10:02
  • Not necessarily shared, it can be ssh. See: docs.opennebula.org/4.14/administration/storage/… Apr 11, 2016 at 10:57
  • Suppose there was a failure and the host poweroff. The data will not be replicated to another host? My task is to make the system fault tolerant. Can I use any alternatives? With minimal points of failure. Apr 11, 2016 at 11:31

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .