Service customization parameters

V-Kube-Service provides a container service with some customizable parameters called service options, including:

• portSpecification

• persistStorage

• region

• resourceUnit

portSpecification

Port specification provides the user with two options for exposing ports for user container services:

• One is to specify the User Specified Service Port. Here, the user specifies the port number exposed on the host for his container service. The selection is not allowed to conflict with the existing port number.
• The second is System Random Map Port. Here, the user chooses to let the VKube system randomly select the port number exposed on the host for the user's container service.

persistStorage

If you select to have persistStorage, VKube will provide persistent storage services in one of the directory paths of the container service, i.e., the data in that directory will be retained even after the service is restarted.

region

VKube currently supports four regions region, including North America, Europe, Tokyo, and Hong Kong, where you can deploy your service.

resourceUnit

resourceUnit indicates the resource unit options you can choose from:

• 1-unit-resource
• 2-unit-resource
• 4-unit-resource
• 8-unit-resource
• 16-unit-resource

There are descriptions about resourceUnit resource group concept:

How resource units are calculated

• 1 Resource-Unit = 1 host port(s) and 1 Gi pod memory
• 2 Resource-Units = 2 host ports and 2 Gi pod memory
• 4 Resource-Units = 4 host ports and 4 Gi pod memory
• 8 Resource-Units = 8 host ports and 8 Gi pod memory
• 16 Resource-Units = 16 host ports and 16 Gi pod memory

Port

With purchased ports and selected User Specified Service Port (or System Random Map Port) when you place an order, you can self-define machine host ports for port forwarding. The number of ports used should be less than or equal to the number of purchased ports.

Memory

The purchased memory is the max amount of memory that can be allocated to the user's service. The memory unit is Gi. When a service deployed by a user exceeds the memory limit, it is possible to trigger memory kill to restart the user's service.

CPU

CPU usage limitation, calculated by the following equation: x: purchased memory size m: total memory of resource node y: available cpu resource for user n: total cpu resource of resource node

y = (x / m) * n

Bandwidth

The network throughput of a pod can't exceed 10Mi * x by default.

PersistStorage

If the user select YES for persistStorage, the VKube system will assign x * 8Gi persistent storage, which x depends on x-Resource-Unit.