Nodes section is used to define Fog resources.
A fog infrastructure is composed of machines with various resource capabilities from low processing power machines,
like Raspberries, to public cloud VMs, like Amazon’s EC2.
To capture this heterogeneity
section allows users to describe the resource characteristics of a physical or
virtual host, including properties for the
Users can determine multiple different Node profiles, which can be used in multiple Fog instances at the instantiation phase.
Following, we demonstrate how a raspberry pi 3b can be described:
nodes: - name: raspberry_pi_3_B capabilities: processor: cores: 4 clock_speed: 1400 memory: 1G disk: type: SDcard size: 8G read: 95MB/s write: 90MB/s
Name is the identifier of the node profile. Users use
name to specify which fog node has the specific node capabilities.
Processor characterizes the CPU capabilities of the current node profile. Specifically, users defines the
that is the number of CPU’s cores and the
clock_speed that is the CPU frequency.
By default, the units of
clock_speed is the megahertz and the user should not define them.
After the deployment, Fogify translates the latter into the
cgroup constraints, dedicated for each running container.
Similarly with processor tag,
memory specifies the amount of RAM that a Fog node is equipped with.
The available units of memory property is
G for gigabyte and
M for megabyte.
Disk defines the node’s characteristics related to storage properties.
disk users can set the
type of storage device (possible values are:
size of storage in gigabyte (
G) or megabyte (
write speed in