As an administrator, you can monitor the servers allocated to tasks or workflows created by users.
This section explains how to monitor your servers and workflows within Rubiscape.

Modes in Rubiscape

Rubiscape platform works in two modes -

  • Scaled Mode
  • Non-Scaled Mode

At the time of deployment, one of the above options can be chosen. This choice is generally based on the amount of data to be processed and the computation power required to process that data.

Scaled Mode

Multiple servers are utilized to perform a specific task in the Scaled Mode based on computation requirements.
There are one main server and other worker servers, which are spawned as required. If a user is performing complex tasks, they utilize the power from these multiple servers. When the server receives a request for workflow execution, workers are spawned to complete the task. When those servers come online, they are listed on the Server Monitoring page. After the execution is complete, the servers go offline.
Thus, the servers come online only for the duration of task execution. There are zero workers when the server is not doing any computation and no workflow runs.
This process is transparent to the user and hence is also called auto-scaling.

(info)
Notes

Scaled mode is recommended

  • For optimized efficiency and effective cost management
  • When the amount of data to be processed is not fixed.

Non-Scaled Mode

Unlike the scaled mode, in the Non-Scaled Mode, a fixed server computation power is utilized for performing user tasks.
According to the changing requirement, the facility to utilize more or fewer servers is not available in this mode. The amount of data and processing speed are restricted to the single server's computational power and speed.

(info)
Notes

  • Non-scaled mode is recommended when the amount of data to be processed is fixed.
  • If your organization uses the Scaled Mode, you can monitor the servers and workers.

Elements of Server

The default Server Monitoring page is shown below. It helps to understand the current status of the server execution.

The table below describes the various elements of the Server Monitoring page. Use this page to view a summary of the following Rubiscape Microservices and their usage.

Field/Icon

Description

  • Execution Server Engine KPI Card
  • See the currently active total execution server engines (workers)

  • Concurrent Process Capability KPI Card
  • See the total concurrent process capability of all the execution server engines (workers)

  • Active Tasks KPI Card
  • See the total number of active tasks of all the execution server engines (workers)

Execution Server Engine

  • Displays a list of the execution server engines (workers) employed for task execution.
  • Only the active workers are listed here.
  • The list changes depending on whether the worker is online or offline.

Execution Server Engine Host

  • Displays the name of the host where the server is running.
  • If the server runs on the same machine, it shows localhost.

Concurrency

  • Displays the number of tasks a server can process at a time.
  • For example, a concurrency of ten (10) indicates that the server can run ten nodes simultaneously.
  • The collective concurrency is displayed in the Concurrent Process Capability KPI card.

Active Tasks

  • Displays the number of tasks currently active on the server.
  • This number is always less than or equal to the individual worker concurrency.
  • The active collective tasks are displayed in the Active Tasks KPI card.

Table of Contents