OpenNebula front-end spawns the following processes when it starts: ( Note that front-end should always be started as oneadmin user using the command sudo -u oneadmin one start )
1) OpenNebula Daemon ( oned ) - Responsible for handling all incoming requests ( either from CLI or from API ). Talks to other processes whenever required.
2) Scheduler ( mm_sched ) - It does match making to find a suitable host ( amongst the hosts part of compute cluster ) for bringing up virtual machines.
3) Information Manager ( one_im_ssh.rb ) - Collects resource availability/utilization information for hosts/VMs respectively. ( Resources include CPU and Memory)
4) Transfer Manager ( one_tm.rb ) - Responsbile for Image Managemnt ( Clone Image, Delete Image, etc. )
5) Virtual Machine Manager ( one_vmm_kvm.rb )- Acts as interface to the underlying Hypervisor. ( All operations to be performed on Virtual Machines go through this interface )
6) Hook Manager ( one_hm.rb ) - Responsible for executing Virtual Machine Hooks. ( Hooks are programs which are automatically triggered on VM state changes. They must be configured prior to starting front-end )
OpenNebula front-end can be configured to a great extent by modifying the contents of file /etc/one/oned.conf. Consider a sample configuration below:
################################################################################
HOST_MONITORING_INTERVAL=10 # Used by Information manager to decide the frequency at which resource availability details have to be collected for hosts
VM_POLLING_INTERVAL=10 # Used by Information manager to decide the frequency at which resource utilization details have to be collected for VMs
VM_DIR=/mnt/onenfs/ # Should be shared across all hosts in compute cluster. Contains Disk Images required for booting VMs
PORT=2633 # All supported API calls are converted to XML-RPC calls. front-end runs an XML-RPC Server on this port to handle these calls
DEBUG_LEVEL=3 # DEBUG_LEVEL: 0 = ERROR, 1 = WARNING, 2 = INFO, 3 = DEBUG
NETWORK_SIZE = 254 # Default size for Virtual Networks ( applicable while using onevnet )
MAC_PREFIX = "00:03" # Default MAC prefix to use while generating MAC Address from IP Address ( applicable while using onevnet )
# The following configuration supports KVM hypervisor. Note that the executables one_im_ssh one_vmm_kvm one_tm and one_hm can be found in /usr/lib/one/mads/
IM_MAD = [
name = "im_kvm",
executable = "one_im_ssh",
arguments = "im_kvm/im_kvm.conf" ]
VM_MAD = [
name = "vmm_kvm",
executable = "one_vmm_kvm",
default = "vmm_kvm/vmm_kvm.conf",
type = "kvm" ]
TM_MAD = [
name = "tm_nfs",
executable = "one_tm",
arguments = "tm_nfs/tm_nfs.conf" ]
HM_MAD = [
executable = "one_hm" ]
1) OpenNebula Daemon ( oned ) - Responsible for handling all incoming requests ( either from CLI or from API ). Talks to other processes whenever required.
2) Scheduler ( mm_sched ) - It does match making to find a suitable host ( amongst the hosts part of compute cluster ) for bringing up virtual machines.
3) Information Manager ( one_im_ssh.rb ) - Collects resource availability/utilization information for hosts/VMs respectively. ( Resources include CPU and Memory)
4) Transfer Manager ( one_tm.rb ) - Responsbile for Image Managemnt ( Clone Image, Delete Image, etc. )
5) Virtual Machine Manager ( one_vmm_kvm.rb )- Acts as interface to the underlying Hypervisor. ( All operations to be performed on Virtual Machines go through this interface )
6) Hook Manager ( one_hm.rb ) - Responsible for executing Virtual Machine Hooks. ( Hooks are programs which are automatically triggered on VM state changes. They must be configured prior to starting front-end )
OpenNebula front-end can be configured to a great extent by modifying the contents of file /etc/one/oned.conf. Consider a sample configuration below:
################################################################################
HOST_MONITORING_INTERVAL=10 # Used by Information manager to decide the frequency at which resource availability details have to be collected for hosts
VM_POLLING_INTERVAL=10 # Used by Information manager to decide the frequency at which resource utilization details have to be collected for VMs
VM_DIR=/mnt/onenfs/ # Should be shared across all hosts in compute cluster. Contains Disk Images required for booting VMs
PORT=2633 # All supported API calls are converted to XML-RPC calls. front-end runs an XML-RPC Server on this port to handle these calls
DEBUG_LEVEL=3 # DEBUG_LEVEL: 0 = ERROR, 1 = WARNING, 2 = INFO, 3 = DEBUG
NETWORK_SIZE = 254 # Default size for Virtual Networks ( applicable while using onevnet )
MAC_PREFIX = "00:03" # Default MAC prefix to use while generating MAC Address from IP Address ( applicable while using onevnet )
# The following configuration supports KVM hypervisor. Note that the executables one_im_ssh one_vmm_kvm one_tm and one_hm can be found in /usr/lib/one/mads/
IM_MAD = [
name = "im_kvm",
executable = "one_im_ssh",
arguments = "im_kvm/im_kvm.conf" ]
VM_MAD = [
name = "vmm_kvm",
executable = "one_vmm_kvm",
default = "vmm_kvm/vmm_kvm.conf",
type = "kvm" ]
TM_MAD = [
name = "tm_nfs",
executable = "one_tm",
arguments = "tm_nfs/tm_nfs.conf" ]
HM_MAD = [
executable = "one_hm" ]
