Experiments
Inject failures and observe the impact on your services.
24-Hour Experiment Summary
Successful vs. failed runs over the last 24 hours
80% pass rate
Successful: 4 (80%)Failed: 1 (20%)
Successful (24h)
4
Failed (24h)
1
Total Runs (24h)
5
Pod Failure
Terminate random pods
api-service
Network Latency
Add network delay
payment-service
CPU Stress
Simulate high CPU load
worker
Disk I/O
Increase disk latency
db-service
Experiment Results
| Experiment | Target | Status | Impact | Recovery Time | Started At | Details | |
|---|---|---|---|---|---|---|---|
| Pod Failure | api-service | Completed | Error rate ↑ 12% | 42s | Oct 10, 14:32 | ||
| Network Latency | payment-service | Completed | Latency ↑ 220ms | 1m 18s | Oct 10, 12:10 | ||
| CPU Stress | worker | Failed | Error rate ↑ 45% | N/A | Oct 10, 10:05 | ||
| Disk I/O | db-service | Completed | Latency ↑ 120ms | 55s | Oct 10, 08:21 | ||
| Network Latency | api-service | Completed | Latency ↑ 195ms | 48s | Oct 09, 16:40 | ||
| Pod Failure | api-service | Completed | Error rate ↑ 9% | 34s | Oct 08, 11:15 |
Service Latency
Pod Failure · api-service (ms)
BeforeDuring Chaos
Error Rate
Pod Failure · api-service (%)
BeforeDuring Chaos