HEAL’s Capacity Forecasting feature helps you predict resource usage trends and future infrastructure requirements based on historical behavior. It supports both manual forecasting and automatic forecasting using Machine Learning Engine (MLE). 

Capacity Trend 

Use the Capacity Trend page to view the historical behavior of services and their usage trends. 

1. Go to Capacity Planning  → Capacity Trend. 

2. Select your Application, Service, and Duration. 

3. View transaction trends and behavior metrics like:

• • Heap Memory Usage 
  • Non-Heap Memory Usage 
  • CPU Utilization 
  • Disk I/O 

Tip: Hover over any point in the graph to see metric values, transaction details, and timestamps. 

Behaviour Metric 

To view the behaviour KPIs, under the memory section, click on Heap Memory Usage. 

The graph below with the events bucketing is displayed.

Hover over the Behaviour KPI graph, to know the detailed information at a specific point/duration. 

Based on the instances value, the bucketing data and frequency data is displayed. |

Generating Capacity Forecast 

Use this when you want full control overgrowth assumptions. 

1. Go to Capacity Forecasting. 
2. Select Generate Capacity Forecast for this Workload Mix to generate a Capacity Forecast for a particular week based on the following:
   • Select For a Week: Generates the list of top “taxing” transactions for the selected week. 
   • Select For a Capacity Metric: Generates the list of top transactions that cause the most stress on the selected capacity KPI.
3. Based on the selection, the following data are obtained: 

• • The system-calculated growth trend for each transaction is depicted by the sparkline. 
  • Based on the current transaction volume and the projected percentage increase in the transaction volume, we can calculate the expected peak throughput. 

4. Select the Capacity Metric of interest, as shown below.

5. The growth plan can be calculated to:

• • Any specific transaction. 
  • Overall growth plan for all transactions. 

You can specify a percentage increase in transaction volume and use the slider to tweak the growth percentage. 

6. Select Generate Capacity Forecast for this Growth Plan, to load the Capacity Forecast heat map at the bottom of the screen. 

7. Hover over the metric in the heat map, to display the following: 

• • • Name of the instance 
    • Name of the metric 
    • Current value of the metric 
    • The projected value of the metric 

Using MLE for Auto-Forecasting 

HEAL’s Machine Learning Engine can auto-calculate the growth percentage without manual inputs. 

Steps: 

1. Enable the MLE Detection toggle. 
2. Select only Duration. 
3. Click Apply. 
4. MLE automatically: 

• • Predicts growth for each transaction 
  • Estimates future throughput 
  • Loads the capacity forecast heatmap 

 MLE Formula: 

Projected Growth % = ((Projected Throughput – Current Throughput) * 100) / Current Throughput 

Was this article helpful?

0 0

Send Feedback

Please Share Your Feedback

How Can We Improve This Article?

Name

Email

Submit