Capacity planning

Capacity planning#

Through Scrutinizer, Plixer One Enterprise can leverage the capabilities of the ML Engine to generate forecasts of future network activity for capacity planning:

  • Apply machine learning techniques to create dynamic baselines for network behavior

  • Extend any Scrutinizer report into the future to forecast trends and predict changes to network activity

  • Use AI-/ML-driven data analysis to predict VPN trends, proactively plan capacity, and align investments with business needs

  • Gain visibility into encrypted VPN tunnels to detect threats

  • Gain visibility into address pool utilization and trend its usage

  • Associate users, devices, and applications with the consumption of bandwidth

Overview#

Scrutinizer includes multiple tools and functions that can enhance a network team’s capacity planning capabilities.

Traffic/behavior baselining#

Using collected flow data, the Plixer ML Engine is able to create dynamic machine learning models of baseline network behavior.

Plixer One Enterprise can use these models to deliver additional capacity planning insights in two ways:

  • Alarms for behavioral deviations that exceed a certain threshold (based on the configured sensitivity) using the Plixer Network Intelligence Anomaly policy

  • Activity/deviation monitoring via Behavior tab when drilling into individual hosts from the Explore > Entities > Hosts view.

Reports#

Scrutinizer’s customizable reports are designed to help teams get to the bottom of any inquiry.

For capacity planning, they can be used to investigate traffic saturation on top interfaces and help determine whether additional provisioning will be required.

Forecasts#

Forecasting is a Plixer One Enterprise feature that allows users to create forecasts of future network activity.

A forecast can be generated from any saved report and will comprise projections for the traffic included by the report configuration (e.g., devices, filters, etc.). This gives teams the ability to define the exact network activity to be forecasted as part of capacity planning.

HD utilization projections#

On the Admin > Resources > System Performance page, clicking on a collector opens a view showing predicted HD utilization based on the current data retention settings. These projections can be used to ensure that sufficient disk space is always available to meet historical data storage needs.

Workflows#

The following workflow(s) show how teams can leverage Plixer One Enterprise functions to enhance their capacity planning capabilities:

Forecasting and meeting business needs

The network team is asked to predict how long an organization’s current infrastructure will continue to support their business needs. To visualize trends in network growth, they create report configurations for various aspects of the environment and use them to create ML-driven forecasts in Scrutinizer.

Workflow

Because Scrutinizer forecasts are based on reports, the environment’s current capabilities should be split up into separate capacities, such as:

  • WAN usage

  • VPN traffic

  • Subnet-to-subnet patterns

  • BGP traffic

  • Core router saturation

  • Critical application latency

From there, one or more report configurations should be created and saved for each capacity. These reports can then be used to generate forecasts that will show emerging utilization trends. At the same time, any latency problems discovered may also indicate potential capacity issues that need to be addressed, depending on their frequency and degree of deviation from the baseline.

Contents