Customer Impact / Cross-Cluster Anomaly

Identify shared infrastructure problems by correlating similar behavior across independent Qumulo clusters.

Unacknowledged

Risk score

56/ 100High

+24 vs. 24h ago

Status

High risk

Confidence: 72%

Estimated lead time

Observed now

Observed in current data

Risk score trend

Correlated latencyqumulo

4 independent clusters show the same latency rise across East US 2

Indicators 1 + 3azure

Platform event and storage degradation share the same zone

What it is: A 0–100 estimate of how likely Customer Impact (Cross-Cluster) are to cause customer impact soon. Higher means more signals are moving away from normal at the same time.
How it is calculated: Each signal is compared with its own baseline (same hour, last 14 days). The further it is above baseline, the more points it adds, up to a fixed weight per signal. The points add up to the score, capped at 100. Method: topology-based correlation + blast-radius calculation.
This score: Increased storage latency +18, Higher API error rate +12, Network RTT to storage +11, other signals +15 = 56 (High), +24 vs. 24h ago.; 0–25 Normal26–50 Elevated51–75 High76–100 Critical
What it tells you: The risk is already visible in current data. Start with Increased storage latency, which adds the most points. Confidence 72% reflects how many expected data feeds reported.

Score breakdown56 / 100

Signal	Current	Baseline	Points	Impact
Increased storage latency	4 clusters	0	+18	high
Higher API error rate	3 clusters	0	+12	medium
Network RTT to storage	4 clusters	1	+11	medium
Read/write throttling	3 clusters	0	+10	medium
Cross-AZ packet loss	2 clusters	0	+5	low

Resource	Type	Region / AZ	Current Risk
rg-anq-prod-eus2-c0142	Qumulo Cluster	East US 2 / 1	86
rg-anq-prod-eus2-c0217	Qumulo Cluster	East US 2 / 1	72
rg-anq-prod-eus2-c0388	Qumulo Cluster	East US 2 / 2	64
rg-anq-prod-eus2-c0451	Qumulo Cluster	East US 2 / 1	58

1Investigate a possible shared dependency across East US 2 before handling these as separate customer tickets.
2Open a proactive case with Microsoft citing the correlated clusters.
3Prepare communication for the 4 exposed clusters if impact is confirmed.