Synology SA3400 Virtual Machine Manager s White Paper - Page 18

Running Regular or High-Availability Clusters, Synology VMM API, a. Computing nodes

Page 18 highlights

connection. Access through VMM on Synology NAS is also required to boot machines that are powered off, install and update operating systems, change BIOS settings, or guarantee successful reboots. Synology VMM API The Synology VMM API allows you to perform some of these routine management operations without accessing the DSM user interface. For more information on API requests and syntax, see the Synology VMM API Guide. 6. Running Regular or High-Availability Clusters Clusters allow you to boost efficiency and availability by separate computing and storage nodes, migrating machines between nodes, and creating failover mechanisms. The hosts in a cluster have the ability to monitor each other. Setting up a VMM virtualization cluster has consequences for network infrastructure, computing power, and storage performance, which are reflected in our recommendations for cluster network configurations and cluster nodes. Providing nodes with the right resources is essential to creating an optimal cluster. Ideal clusters feature different nodes that are optimized for their specific roles. The below sections explore the best ways to optimize your Synology NAS and network for cluster use. a. Computing nodes Computing nodes run your VMs, and handle communication with the local network and the internet. Their configuration determines the maximum number and speed of your VMs. Higher-spec CPUs, RAM, and network interface cards are recommended on compute nodes. Adding the maximum amount of RAM to computing nodes and using multiple network interfaces boosts performance. Using all-SSD storage likewise speeds up VM operations. Performance comparisons can be found under Comparing Configurations. An essential feature of any cluster (VMM High Availability and regular) is the ability to migrate VMs to another node ahead of expected downtime or to balance workloads. Ideally, two or more Synology NAS with very similar capabilities should be configured as computing nodes. Less demanding services can also temporarily be migrated to weaker nodes, such as storage nodes. At a minimum, this requires that the nodes share the same CPU architecture. We maintain a comprehensive list of CPU and RAM configurations in Support Center. b. Processor compatibility Different CPUs come with different instruction set architecture features that significantly boost performance for certain processes. Operating systems and applications detect such features when booting or launching, and adjust their operations to the host CPU. Moving a running VM (live migration) or a VM in a saved state (failover) to a node with a different processor architecture without rebooting can cause machines and applications to fail or crash. CPU compatibility mode is a functionality that allows live migration between two hosts with different processors. 16 Configuration Guide

  • 1
  • 2
  • 3
  • 4
  • 5
  • 6
  • 7
  • 8
  • 9
  • 10
  • 11
  • 12
  • 13
  • 14
  • 15
  • 16
  • 17
  • 18
  • 19
  • 20
  • 21
  • 22
  • 23
  • 24
  • 25
  • 26
  • 27
  • 28
  • 29
  • 30

16
Configuration Guide
connection. Access through VMM on Synology NAS is also required to boot machines that
are powered off, install and update operating systems, change BIOS settings, or guarantee
successful reboots.
Synology VMM API
The Synology VMM API allows you to perform some of these routine management operations
without accessing the DSM user interface. For more information on API requests and syntax,
see the
Synology VMM API Guide.
6. Running Regular or High-Availability Clusters
Clusters allow you to boost efficiency and availability by separate computing and storage
nodes, migrating machines between nodes, and creating failover mechanisms. The hosts in a
cluster have the ability to monitor each other.
Setting up a VMM virtualization cluster has consequences for network infrastructure,
computing power, and storage performance, which are reflected in our recommendations
for cluster network configurations and cluster nodes.
Providing nodes with the right resources is essential to creating an optimal cluster. Ideal
clusters feature different nodes that are optimized for their specific roles. The below sections
explore the best ways to optimize your Synology NAS and network for cluster use.
a. Computing nodes
Computing nodes run your VMs, and handle communication with the local network and the
internet. Their configuration determines the maximum number and speed of your VMs.
Higher-spec CPUs, RAM, and network interface cards are recommended on compute nodes.
Adding the maximum amount of RAM to computing nodes and using multiple network
interfaces boosts performance. Using all-SSD storage likewise speeds up VM operations.
Performance comparisons can be found under
Comparing Configurations
.
An essential feature of any cluster (VMM High Availability and regular) is the ability to migrate
VMs to another node ahead of expected downtime or to balance workloads. Ideally, two or
more Synology NAS with very similar capabilities should be configured as computing nodes.
Less demanding services can also temporarily be migrated to weaker nodes, such as storage
nodes. At a minimum, this requires that the nodes share the same CPU architecture. We
maintain a
comprehensive list
of CPU and RAM configurations in Support Center.
b. Processor compatibility
Different CPUs come with different instruction set architecture features that significantly
boost performance for certain processes. Operating systems and applications detect such
features when booting or launching, and adjust their operations to the host CPU.
Moving a running VM (live migration) or a VM in a saved state (failover) to a node with a
different processor architecture without rebooting can cause machines and applications to
fail or crash. CPU compatibility mode is a functionality that allows live migration between two
hosts with different processors.