|
|
|
VM Mobility |
|
|
|
Live Migration of VMs
Details
|
Yes
Each cluster is configured with a minimal CPU type that all hosts in that cluster must support (you specify the CPU type in the RHEV-M GUI when creating the cluster). Guests running on hosts within the cluster all run on this CPU type, ensuring that every guest can be live migrated to any host within the cluster. This cannot be changed after creation without significant disruption. All hosts in a cluster must run the same CPU type (Intel or AMD).
|
Yes - Live Migration
There is also a possibility to use SSL to encryption the Live Migration. More info on: http://docs.oracle.com/cd/E50245_01/index.html
|
Yes vMotion
- Cross vSwitch vMotion Only
vMotion
- Cross-vCenter vMotion with mixed-version (new)
- Cross vSwitch vMotion (all versions)
- Cross vCenter vMotion (n/a for Standard)
- Long Distance vMotion (n/a for Standard)
- Cross Cloud vMotion (n/a for Standard)
- Encryption vMotion (n/a for Standard)
|
|
|
Migration Compatibility
Details
|
Yes
|
Yes
On the same Cluster/Server Pool it is possible to separate the Cluster/Server Pool Members based on the CPU Family to guarantee the success of Live Migration between those Servers but there is no possibility of grouping different CPUs generations by masking out incompatible functions using CPU Masking. More info on: http://docs.oracle.com/cd/E50245_01/index.html
|
Yes (EVC)
EVC can be enabled Per-VM (new)
Enhanced vMotion Compatibility - enabled on vCenter cluster-level, utilizes Intel FlexMigration or AMD-V Extended Migration functionality available with most newer CPUs (but cannot migrate between Intel and AMD), Details here: http://kb.vmware.com/kb/1005764
|
|
|
|
Yes - Built-in (CPU\Memory) and plugable scheduler
NEW
A policy engine determines the specific host on which a virtual machine runs. The policy engine decides which server will host the next virtual machine based on whether load balancing criteria have been defined, and which policy is being used for that cluster. RHEV-M will use live migration to move virtual machines around the cluster as required.
A scheduler handles virtual machine placement, allowing users to create new scheduling policies, and also write their own logic in Python and include it in a policy.
- The scheduler serves scheduling requests for running or migrating virtual machines according to a policy.
- The scheduling policy also includes load balancing functionality.
- Scheduling is performed by applying hard constraints and soft constraints to get the optimal host for that request at a given point of time
- The infrastructure allowing users to extend the new scheduler, is based on a service called ovirt-scheduler-proxy. The services purpose is for RHEV admins to extend the scheduling process with custom python filters, weight functions and load balancing modules.
- Every cluster has a scheduling policy. Administrators can create their own policies or use the built-in policies which were extended to support new capabilities such as shutting down servers for power saving policy.
The load balancing process runs once every minute for each cluster in a data center. You can disable automatic migration for individual vm or pin them to specific hosts.
You can choose to set the policy as either even distribution or power saving, but NOT both.
|
Yes - Maintenance Mode
It is possible to execute the Maintenance Mode and It is also possible to Lock some Servers to not receive this Offload from the Server that is entering in this State. More info on: http://docs.oracle.com/cd/E50245_01/index.html
|
Yes
Maintenance mode is a core feature to prepare a host to be shut down safely.
vSphere Quick Boot is a new innovation that restarts the ESXi hypervisor without rebooting the physical host. (new)
|
|
|
Automated Live Migration
Details
|
Yes
When Power saving is enabled in a cluster it distributes the load in a way that consolidates virtual machines on a subset of available hosts. This enables surplus hosts that are not in use to be powered down, saving power. You can set the threasholds in the RHEV-M GUI to specify the Minimum Service Level a host is permitted to have.
You must also specify the time interval in minutes that a host is permitted to run below the minimum service level before remaining virtual machines are migrated to other hosts in the cluster - as long as the maximum service level set also permits this.
|
Yes
Oracle VM provides a DRS feature for the following constraints: (CPU-CPU, N:Network i/o) during VMs runtime and the following constraints: (CPU-CPU, N:Network i/o, MEM:Memory) during VMs startup. More info on: http://docs.oracle.com/cd/E50245_01/index.html
|
No
VM Distribution: Enforce an even distribution of VMs.
Memory Metric for Load Balancing: DRS uses Active memory + 25% as its primary metric.
CPU over-commitment: This is an option to enforce a maximum vCPU:pCPU ratios in the cluster.
Network Aware DRS - system look at the network bandwidth on the host when considering making migration recommendations.
Storage IO Control configuration is now performed using Storage Policies and IO limits enforced using vSphere APIs for IO Filtering (VAIO).
Storage Based Policy Management (SPBM) framework, administrators can define different policies with different IO limits, and then assign VMs to those policies.
|
|
|
|
Yes
NEW
Storage Live Migration is supported and allows migration of virtual machine disks to different storage devices while the virtual machine is running. There is also an option to move an entire storage domain between datacenters or even between setups.
|
Yes
Oracle VM provides a DPM feature that works with IPMI and Wake Up On Lan. More info on: http://docs.oracle.com/cd/E50245_01/index.html
|
No
Distributed Power Management - enables to consolidate virtual machines onto fewer hosts and power down unused capacity - reducing power and cooling. This can be fully automated where servers are powered off when not needed and powered back on when workload increases.
|
|
|
Storage Migration
Details
|
200 hosts/cluster
That is the supported maximum number of hosts per RHEV datacenter and also per cluster (the theoretical KVM limit is higher).
|
Yes
Oracle VM provides a capability to Migrate the VM Disks to different Storage Repositories (not live). More info on: http://docs.oracle.com/cd/E50245_01/index.html
|
Yes (Live Storage vMotion)
Storage vMotion allows to perform live migration of virtual machine disk files (e.g. across heterogeneous storage arrays) without vm downtime. Storage DRS handles initial vmdk placement and gives migration recommendations to avoid I/O and space utilization bottlenecks on the datastores in the cluster. The migration is performed using storage vMotion.
|
|
|
|
HA/DR |
|
|
|
|
Yes
High availability is an integrated feature of RHEV and allows for virtual machines to be restarted in case of a host failure.
HA has to be enabled on a virtual machine level. You can specify levels of priority for the vm (e.g. if resources are restrained only high priority vm are being restarted). Hosts that run highly available vm have to be configured for power management (to ensure accurate fencing in case of host failure).
Fencing Details: When a host becomes non-responsive it potentially retains the lock on the virtual disk images for virtual machines it is running. Attempting to start a virtual machine on a second host could cause data corruption. Fencing allows RHEV-M to safely release the lock (using a fence agent that communicates with the power management card of the host) to confirm that a problem host has truly been rebooted.
RHEV-M gives a non-responsive host a grace period of 30 seconds before any action is taken in order to allow the host to recover from any temporary errors.
Note: The RHEV-M manager needs to be running for HA to function (unlike e.g. VMware HA or Hyper-V HA that do not rely on vCenter / VMM for the failover capability), also HA can not be enabled on the cluster level.
|
32 Hosts in Cluster/Pool
Oracle VM can have up to 32 Servers on each Clustered Server Pool and up to 64 Servers on each Unclustered Server Pool. More info on: http://docs.oracle.com/cd/E50245_01/index.html
|
max 64 nodes / 8000 vm per cluster
Up to 64 nodes can be in a DRS/HA cluster, with a maximum of 8000 vm/cluster
|
|
|
Integrated HA (Restart vm)
Details
|
Yes (HA, WatchDog)
RHEV supports watchdog device for linux guests that restarts virtual machines in case of OS failure. High availability (in addition to monitoring physical hosts) also monitors all virtual machines, so if the virtual machines operating system crashes, a signal is sent to automatically restart the virtual machine, but this is with host change.
|
Yes - Clustered Pool
Oracle VM will restart HA-enabled Guests on remaining Hosts in case of Host failure. More info on: http://docs.oracle.com/cd/E50245_01/index.html
|
Yes (VMware HA)
vSphere 6.5 Proactive HA detect hw condition evacuate host before failure (plugin provided OEM vendors)
Quarantine mode - host is placed in quarantine mode if it is considered in degraded state
Simplified vSphere HA Admission Control - 'Percentage of Cluster Resources' admission control policy
vSphere 6.0
- Support for Virtual Volumes – With Virtual Volumes a new type of storage entity is introduced in vSphere 6.0.
- VM Component Protection – This allows HA to respond to a scenario where the connection to the virtual machine’s datastore is impacted temporarily or permanently.
“Response for Datastore with All Paths Down”
“Response for Datastore with Permanent Device Loss”
- Increased scale – Cluster limit has grown from 32 to 64 hosts and to a max of 8000 VMs per cluster
- Registration of “HA Disabled” VMs on hosts after failure.
VMware HA restarts virtual machines according to defined restart priorities and monitors capacity needs required for defined level of failover.
vSphere HA in vSphere 5.5 has been enhanced to conform with virtual machine-virtual machine anti-affinity rules. Application availability is maintained by controlling the placement of virtual machines recovered by vSphere HA without migration. This capability is configured as an advanced option in vSphere 5.5.
vSphere 5.5 also improved the support for virtual Microsoft Failover Clustering (cluster nodes in virtual machines) - note that this functionality is independent of VMware HA and requires the appropriate Microsoft OS license and configuration of a Microsoft Failover Cluster. Microsoft clusters running on vSphere 5.5 now support Microsoft Windows Server 2012, round-robin path policy for shared storage, and iSCSI and Fibre Channel over Ethernet (FCoE) for shared storage.
While not obvious to the user - with vSphere 5, HA has been re-written from ground-up, greatly reducing configuration and failover time. It now uses a one master - all other slaves concept. HA now also uses storage path monitoring to determine host health and state (e.g. useful for stretched cluster configurations).
|
|
|
Automatic VM Reset
Details
|
No
There is no live lock-step mirroring support in RHEV - although the theoretical capability is available in KVM. Red Hat tends to points out that the limitations around this technology (inability to take e.g. snap shots, perform a live storage migrate, limited guest vCPU support, high bandwidth/processing requirements) can make it inappropriate for enterprise implementation.
|
Yes
Oracle VM will restart Guests on the same Host due to OS Failures. More info on: http://docs.oracle.com/cd/E50245_01/index.html
|
Yes (VMware HA)
vSphere 6.5 - Orchestrated Restart - VMware has enforced the VM to VM dependency chain, for a multi-tier application installed across multiple VMs.
Uses heartbeat monitoring to reset unresponsive virtual machines.
|
|
|
VM Lockstep Protection
Details
|
No (native);
Yes (with Vendor Add-On: Red Hat Cluster Suite)
There is no integrated application level monitoring or restart of services/vm in case of application failures. RHEV supports watchdogs and HA.
This is possible using Red Hat Cluster Suite. This is a Fee-based Add-On.
|
No
|
Yes (Fault Tolerance) 2 vCPUs.
NEW
vSphere 6.7 now supports 8 vCPUs and 128 GB RAM per VM.
vSphere 6.5 FT has more integration with DRS and enhanced Network (lower the network latency)
Fault Tolerance brings continuous availability protection for VMs with up to 4 vCPUs in Enterprise Plus and Standard is 2 vCPUs.
Uses a shadow secondary virtual machine to run in lock-step with primary virtual machine to provide zero downtime protection in case of host failure.
|
|
|
Application/Service HA
Details
|
No - See Details
There is no natively provided Site Failover capability in RHEV. Red Hat does provide the tools needed to provide a disaster recovery solution.
This is possible via 3rd party partners integration (such as Veritas, Acronis, SEP, Commvault).
|
No
|
No
vSphere 6.5 Proactive HA detect hw condition evacuate host before failure (plugin provided OEM vendors)
Quarantine mode - host is placed in quarantine mode if it is considered in degraded state
Simplified vSphere HA Admission Control - 'Percentage of Cluster Resources' admission control policy
VMware HA restarts virtual machines according to defined restart priorities and monitors capacity needs required for defined level of failover.
|
|
|
Replication / Site Failover
Details
|
Yes
NEW
You are able to update both RHEV-H or RHEL-H via the management UI. The management sends events on updates pending on the hosts and the manager machine.
Updates can be also managed via Red Hat Satellite.http://red.ht/1Oxs20B
|
Yes
If the Storage is Replicated the whole environment can be orchestrated to be up and running in minutes if the Failover Site already has the Oracle VM up and running pointing to the Replicated Storage with the help of Oracle Site Guard. More info on: http://www.oracle.com/technetwork/server-storage/vm/ovm3-disaster-recovery-1872591.pdf
|
Yes (vSphere Replication)
DR Orchestration (Vendor Add-On: VMware SRM)
vSphere Replication is VMware’s proprietary hypervisor-based replication engine designed to protect running virtual machines from partial or complete site failures by replicating their VMDK disk files.
This version extends support for the 5 minute RPO setting to the following new data stores: VMFS 5, VMFS 6, NFS 4.1, NFS 3, VVOL and VSAN 6.5.
|
