Slow Degradation Detection Challenge

A transaction counter degrades slowly over a 1-hour cycle. Your monitoring agent must detect when performance thresholds are crossed and report the correct severity level.

Transaction Counter

5,151 TPS
WARNING
10,000 TPS 36:22 / 60:00 2,000 TPS
OK
≥ 6,000 TPS
WARNING
4,000 – 5,999 TPS
ALERT
< 4,000 TPS

Verification

This signature confirms the page loaded correctly. Your agent must also read the counter value and report the appropriate severity level.

SINAINSIGHT-DEGRADATION-VERIFIED

Challenge Instructions

Goal: Detect slow performance degradation by monitoring a transaction counter that drops over time, and report the correct severity level.

Degradation cycle: The counter starts at 10,000 TPS and drops linearly to 2,000 TPS over 60 minutes, then resets.

Thresholds:
1. Read the transaction counter from element degradation-counter-value.
2. ≥ 6,000 TPS → OK — performance is normal.
3. 4,000 – 5,999 TPS → WARNING — performance is degrading.
4. < 4,000 TPS → ALERT — performance is critical.

Expected behavior: When your agent runs periodically (e.g. every 5 minutes), it should report level changes as the counter crosses thresholds. Over a full 1-hour cycle, a properly configured monitor will show the transition from OK → WARNING → ALERT → OK (reset).

Technique: Real-world APM systems must detect gradual degradation, not just hard failures. A server returning HTTP 200 with increasing latency or decreasing throughput requires threshold-based monitoring to catch problems before they become outages. This challenge simulates that pattern with a predictable 1-hour cycle.