Load Testing Issues¶

Load test won't start, or failing during execution? This page covers the most common load test problems. Find the symptom that matches what you're seeing.

Load Test Won't Start¶

"Unable to connect to load engine"¶

Symptom: Starting load test fails with "connection refused" to engine.

Most common causes: 1. Cloud engine not started (EC2 instance stopped) 2. Firewall blocking ports 1099/1100 3. Wrong IP address configured for engine 4. Engine startup not complete (takes 2-3 minutes)

Fix: 1. Check engine status - AWS console → verify EC2 instance is running 2. Verify IP address - Engines View → check IP matches EC2 instance 3. Check firewall rules - Must allow TCP ports 1099 and 1100 4. Wait for engine startup - Engines take 2-3 minutes to fully start after EC2 launch

See: Cloud Load Testing for engine configuration.

"Engine startup timeout"¶

Symptom: Load test won't start, engine times out.

Most common causes: 1. Engine didn't respond within timeout period 2. Network issue between Load Tester and engine 3. Engine crashed or overloaded 4. Engine configuration problem

Fix: 1. Check engine logs - SSH into engine, check /var/log/wpi-engine.log 2. Restart engine - Stop and restart EC2 instance 3. Check network connectivity - Ping engine from Load Tester workstation 4. Increase timeout - Tools → Preferences → Engines → Startup Timeout

"Insufficient engine capacity"¶

Symptom: Configuring load test shows error about VU capacity.

Most common causes: 1. Trying to run more VUs than engines can handle 2. Engine instance type too small 3. Not enough engines for target load

Fix: 1. Add more engines - Each engine handles ~500-1000 VUs depending on test complexity 2. Use larger instance types - Switch from t3.medium to c5.xlarge for more capacity 3. Reduce VU count - Start with lower load, add engines as needed

See: Cloud Load Testing for capacity planning.

"Test case upload failed"¶

Symptom: Starting load test fails during test case upload to engines.

Most common causes: 1. Network timeout 2. Test case too large (large datasets) 3. Engine disk full

Fix: 1. Check test case size - Very large datasets can cause upload timeouts 2. Reduce dataset size - Split large datasets across multiple test cases 3. Check engine disk space - SSH to engine, run df -h 4. Retry upload - Temporary network glitches can cause this

Load Test Fails During Execution¶

"Connection pool exhausted"¶

Symptom: Load test running, errors about connection pool.

Most common causes: 1. Test case creates too many concurrent connections 2. TIME_WAIT sockets not closing fast enough 3. OS connection limits reached

Fix: 1. Increase OS connection limits - Engines View → Advanced → Tune OS parameters 2. Add keep-alive headers - Reuse HTTP connections instead of creating new ones 3. Reduce VU ramp-up rate - Slower ramp-up gives connections time to close 4. Check TIME_WAIT sockets - netstat -an | grep TIME_WAIT | wc -l on engine

"Connection refused errors during load test"¶

Symptom: Load test runs but many requests fail with connection refused.

Most common causes: 1. Server can't handle the load (listen queue exhausted) 2. Firewall rate limiting 3. Server reached connection limit 4. Server crashed under load

Fix: 1. Check server health - Is server still responding? Check server logs 2. Reduce load - Lower VU count to find sustainable load level 3. Increase server listen queue - Server OS tuning may be needed 4. Check firewall rules - Some firewalls rate-limit connections

See: Load Testing Basics for load ramping strategies.

"Timeout errors during load test"¶

Symptom: Many requests timing out during load test.

Most common causes: 1. Server performance degraded under load (CPU/memory exhausted) 2. Database locks or slow queries 3. Network latency increased 4. Timeout value too aggressive for server response time

Fix: 1. Check server metrics - CPU, memory, disk I/O - see Server Monitoring 2. Increase timeout values - Test Case Properties → increase request timeout 3. Reduce load - Lower VU count to find where timeouts start 4. Monitor database - Check for slow queries or lock contention

"Rate limiting or throttling errors"¶

Symptom: Server returns 429 "Too Many Requests" or similar throttling errors.

Most common causes: 1. API rate limits exceeded 2. DDoS protection triggered 3. Server-side throttling policy

Fix: 1. Add think time - Slow down request rate per VU 2. Contact server team - Request rate limit increase for load testing 3. Whitelist load engine IPs - Exclude test traffic from rate limiting 4. Distribute load across more IPs - Use more engines with different IPs

"VPN connection issues during load testing"¶

Symptom: Load test fails when testing through VPN.

Most common causes: 1. VPN can't handle the connection volume 2. VPN disconnects under heavy load 3. VPN routes timing out

Fix: 1. Test without VPN first - Isolate whether VPN is the problem 2. Use split tunnel VPN - Only route test traffic through VPN, not all traffic 3. Increase VPN timeout settings 4. Consider direct connection - Test from location that doesn't require VPN

"Out of memory errors during load test"¶

Symptom: Load Tester crashes or shows out of memory errors.

Most common causes: 1. Too many VUs for workstation RAM 2. Test results accumulating too much data 3. Large response payloads

Fix: 1. Increase Java heap - Edit LoadTester.ini, add -Xmx4g (4GB heap) 2. Use more engines, fewer VUs per engine - Distribute load 3. Reduce result detail - Tools → Preferences → Results → reduce sampling rate 4. Upgrade workstation RAM - Large tests require 8+ GB RAM

Load Test Results Problems¶

"Cannot open test results: file corrupted"¶

Symptom: Opening test results shows corruption error.

Most common causes: 1. Disk full during test 2. Load Tester crashed during test 3. Power loss during test

Fix: 1. Try backup file - Load Tester creates .wpt.backup files automatically 2. Check disk space - Ensure drive has space (test results can be large) 3. Export partial results - Right-click → Export → choose format 4. Contact support - We may be able to recover partial data

"Out of memory while generating report"¶

Symptom: Opening load test results or generating reports fails with out of memory.

Most common causes: 1. Test results too large for available RAM 2. Java heap too small

Fix: 1. Increase Java heap - Edit LoadTester.ini, add -Xmx4g (4GB heap) 2. Close other applications - Free up RAM 3. Export subsets - Export results in chunks (first 10 minutes, next 10 minutes) 4. Upgrade RAM - Large tests (hours long, many VUs) require 8+ GB RAM

Error Message Lookup¶

Looking for a specific error message? Use the searchable error catalog:

See: Common Error Messages - Use Ctrl+F to search for the exact error text you're seeing.

Still Stuck?¶

If none of these solutions work:

Check cloud/engine issues: Cloud & Engine Issues for AWS-specific problems
Check server monitoring: Server Monitoring to diagnose server-side bottlenecks
Contact Support: See Getting Support for how to gather diagnostic logs

Most load test failures fall into two buckets: server capacity issues (you need server monitoring to diagnose these) and engine configuration problems (firewall, capacity, network). Knowing which bucket you're in is half the battle.