Home Planning Phase Work Load Model

Work Load Model

Work Load Model is the heart of all the documents which will lead you "How to test that application?" & "What to test that application?"

The Work Load Model (WLM) is the blueprint of user activities. It determines the transaction volumes, pacing, and think times needed to simulate realistic user load on the server.

Purpose

Defines what, how much and how fast to test.

Focus

CBT's, Business flows, No. of Transactions, Load distribution by test type.

Role

Guides scenario design, execution and capacity planning.

Key Consideration

Pacing & Think time calculation to generate Anticipated Load.

Outcome

Realistic user simulation and accurate performance results.

What Does Work Load Model Contain?

  • Critical Business Transactions (CBT's)
  • Business flows for every CBT's
  • Number of transactions per CBT
  • Types of testing (Load, Stress, Endurance, Spike, Soak, Baseline, etc.)
  • Load distribution based on types of testing
  • Pacing and Think time calculation
  • Anticipated load and concurrency

Note

Work load model will assist you to design the scenario.
Input:
Use cases, types of testing, number of users
Output:
Pacing, think time and transaction mix

Workload Modeling Terminology

Critical Business Transactions (CBTs)
Key user flows that consume resources (e.g. Login, Search, Checkout).
Think Time
The delay between actions that simulates human reading/typing speed (typically 3 to 10 seconds).
Pacing
The delay between iterations that controls the transaction rate per hour.

WLM Inputs and Outputs

A workload model converts user transaction counts into simulation parameters.

Inputs Processing Logic Outputs
Peak Concurrent Users Distribute users across CBT paths based on business flow mix
Transaction Mix Ratio (%)
Target Transactions per Hour Apply Little's Law formulas to calculate pacing
Pacing Delay Configuration
(Time between iterations)
Human Interaction Delays Add realistic static/random pause (Think time)
Think Time Settings
(Per transaction/user action)
Business Rules & Constraints Apply peak hours, ramp-up/down, and test type characteristics
Anticipated Load Profile
(Users, TPS, Concurrency)

Formula Guidance

Pacing controls the transaction rate. If pacing is too short, you will hammer the server and exceed targets; if too long, you won't reach the required transactional throughput.

Little's Law:
Concurrency (L) = Throughput (λ) × Response Time (W)
Where,
L = Number of concurrent users
λ = Throughput (Transactions per second)
W = Average response time (seconds)
Pacing Formula:
Pacing (sec) =
Total Test Duration (sec)
Total Iterations
Or
Pacing = 3600 (sec) / Transactions per Hour (TPH)
Adjust pacing with think time to achieve target TPH.

Work Load Model Example (Sample)

CBT ID Critical Business
Transaction		 Business Flow
(High Level)		 Transactions Transaction Mix (%) Think Time (Sec) Pacing (Sec)
Load Stress Spike Soak Baseline Min Max Avg
CBT-01 Login Open App > Login > Dashboard 3 10%10%10%10%10% 375 6.5
CBT-02 Search Search Items > View Results 5 20%20%20%20%20% 264 4.2
CBT-03 Add to Cart Select Item > Add to Cart 4 15%15%15%15%15% 253.5 5.1
CBT-04 Checkout Cart > Checkout > Payment 8 25%25%25%25%25% 385.5 7.0
CBT-05 Logout Logout > End Session 2 5%5%5%5%5% 132 2.8
Total 22 100%100%100%100%100% --- -