NOFire.ai logo
Fast Incident Clarity

Get to root cause in 2 minutes

From alert to fix, with full production context

2 min
Root Cause
70%
Faster MTTR
91%
Accuracy

From alert to resolution in minutes.

Four capabilities that trace every incident from first symptom to root cause and actionable fix.

Instant root cause in Slack

From alert to root cause in 2 minutes with full Grafana context and causal analysis.

# auto-alerts
🔔 1
Grafana
GrafanaBotAPP2:15 PM
FIRING:1MongoDB Atlas Slow Queries prod-mongodb-us-east-1
**Firing**
Value: 5 slow queries (>100ms) detected in last hour
Labels:
- alertname = MongoDB Slow Queries
- cluster = prod-mongodb-us-east-1
- database = userprofiles
- collection = users
Annotations:
COLLSCAN operations: 50-54ms, $collStats operations: 41-52ms
Source: https://cloud.mongodb.com/v2/prod-mongodb-us-east-1/metrics/replicaSet/...
🔥 Grafana v12.0.0Jul 22nd
NOFire AIAPP2:17 PM
Root Cause:
Missing Indexes Triggering Full Collection Scans
Summary:
Investigation of MongoDB Atlas performance degradation reveals missing indexes on userprofiles.users collection. COLLSCAN operations examining 918 documents to return 1 result, causing 50-54ms query latency. accountType and registrationDate fields lack proper indexing. Secondary factor: administrative $collStats operations taking 41-52ms.
Confidence: 92% | Generated by NOFire AI

Causal flow analysis

Visual map from root cause through intermediate effects to final impacts with confidence scores.

CAUSAL FLOW ANALYSIS

Left-to-right flow showing root causes → intermediate effects → final impacts (≥80% confidence only)

Missing Indexes

92%
Root Cause
userprofiles.users

Full Collection Scans

95%
Effect
prod-mongodb-us-east-1

Slow Query Performance

93%
Effect
userprofiles.users

Service Degradation

88%
Impact
user-profile-service

Actionable recommendations

Prioritized fixes from missing indexes to query optimizations with severity levels and investigation paths.

!
NEEDS ATTENTION• 2 issues

EXECUTIVE SUMMARY

MongoDB Atlas cluster prod-mongodb-us-east-1 shows 5 slow queries (>100ms) in last hour. Key findings: COLLSCAN operations on userprofiles.users (50-54ms), aggregate $collStats operations (41-52ms). Performance impact moderate with table scans affecting query efficiency.

Findings & Issues

Collection Scan Operations on userprofiles.users

MEDIUM

slow queries • prod-mongodb-us-east-1

Impact

Inefficient table scans examining 918 documents to return 1 result. Poor query selectivity causing unnecessary resource consumption.

Most Likely Cause

  • •Missing indexes on accountType and registrationDate fields

Administrative $collStats Operations

LOW

slow queries • prod-mongodb-us-east-1

Impact

Monitoring operations taking 41-52ms. Normal administrative overhead but contributing to slow query count.

Most Likely Cause

  • •Large collection statistics requiring time to compute

Top Recommendations

INVESTIGATE (3 OF 4)

  • •Create compound index on userprofiles.users collection for {accountType: 1, registrationDate: 1, status: 1}
  • •Review query patterns in userprofiles.users to optimize filter selectivity
  • •Consider adding single-field indexes on frequently queried fields

MONITOR

  • •Monitor $collStats operations frequency if they become more frequent

Complete incident report

Auto-generated postmortem with root cause, impact summary, and timeline that becomes searchable knowledge.

POSTMORTEM

Severity:SEV-3(Performance Degradation)
Status:Resolved

Root Cause

Missing indexes on accountType and registrationDate fields in userprofiles.users collection, forcing COLLSCAN operations examining 918 documents to return 1 result.

Impact

Query latency increased to 50-54ms for table scans, causing poor query selectivity and unnecessary resource consumption across user-profile-service.

INCIDENT TIMELINE

5 events
14:00QUERY ANOMALYby NOFire AI

MongoDB Atlas slow query detected: COLLSCAN on userprofiles.users taking 50-54ms

14:15ALERT FIRINGby Grafana

Alert: MongoDB Atlas Slow Queries - 5 queries exceeding 100ms threshold

14:20QUERY PATTERNby NOFire AI

Detected pattern: 918 documents examined to return 1 result, indicating poor query selectivity

14:22SECONDARY ISSUEby NOFire AI

Administrative $collStats operations contributing 41-52ms to slow query count

14:25RCA COMPLETEby NOFire AI

Root cause identified: Missing indexes on accountType and registrationDate fields causing full table scans

What fast incident clarity looks like.

For developers

  • Root cause in minutes, not hours of debugging
  • No more guessing or waiting for seniors to diagnose
  • Back to shipping features instead of firefighting

For SRE & platform

  • MTTR reduced from hours to minutes
  • Fewer escalations and war rooms
  • Clear explanations you can trust and act on

For leadership

  • Reduced customer-impacting downtime
  • Reclaimed engineering capacity from incident triage
  • Faster recovery protects revenue and trust

Ready to cut MTTR from hours to minutes?