---
title: "Database Performance Degradation Template"
description: "Technical template for communicating database latency and performance issues to your engineering-focused audience."
author: "openstatus"
publishedAt: "2026-01-19"
category: "template"
faq:
  - question: "How do I decide between technical and user-friendly messaging?"
    answer: "Match your communication to your audience. For developer-focused products, include technical details like p95 latency, query times, and specific database components. For general users, focus on impact: 'slower response times' instead of 'elevated connection pool exhaustion.'"
  - question: "Should I share the root cause of database performance issues?"
    answer: "Yes, once resolved. Technical audiences appreciate transparency and can learn from your incidents. Share root cause, mitigation steps taken, and preventative measures. This builds trust and demonstrates engineering maturity."
  - question: "When should I trigger a database performance incident notification?"
    answer: "Trigger notifications when p95 latency exceeds 2x normal, error rates exceed 1%, connection timeouts occur, or when user reports indicate widespread slowness. Set up automated monitoring alerts to catch these thresholds early."
---

Use this template when experiencing database performance issues, elevated latency, or query slowdowns. Ideal for engineering teams and technical stakeholders.

## When to Use This Template

- Elevated database latency
- Slow query performance
- Connection pool exhaustion
- Database infrastructure issues

## Template Messages

### Investigating

We are experiencing elevated database latency affecting some features. Users may experience slower response times.

Our database team is actively investigating and working to restore normal performance levels.

### Identified

We have identified the root cause. Our team is implementing a fix.

### Monitoring

Database performance improvements have been deployed. Latency is returning to normal levels. We are continuing to monitor.

### Resolved

Database performance has been fully restored. All queries are executing at normal speeds.

## Real-World Examples

### GitHub: "Infrastructure update to data stores"

**Context**: Data store infrastructure changes
**Duration**: ~1.5 hours
**Impact**: 1.8% combined failure rate, peaking at 10%

**What they did well**:
- Quantified impact with specific percentages
- Identified multiple affected services
- Committed to post-incident RCA (Root Cause Analysis)
- Provided precise UTC timestamps

**Sample messaging**: "Infrastructure update to data stores caused 1.8% combined failure rate, peaking at 10% across multiple services Jan 15, 16:40-18:20 UTC."

### Fly.io: "Network instability in IAD region"

**Context**: Regional infrastructure issue
**Approach**: Geographic specificity

**What they did well**:
- Identified specific region (IAD = Ashburn, Virginia)
- Focused on infrastructure layer
- Technical but clear naming

This demonstrates the value of being specific about location when database issues are region-specific.

## Tips for Technical Communication

1. **Include specific metrics** when you have them (p95 latency, error rates, query times)
2. **Name the layer** - application, database, network, storage
3. **Be technical if your audience is technical** - don't oversimplify for engineers
4. **Quantify impact** - "affecting 2% of queries" is better than "some users"
5. **Share root cause** when resolved - technical teams appreciate learning

## Customization Ideas

For **developer audiences**:
```
We are experiencing elevated p95 latency (2.5s vs normal 150ms)
on write operations to our primary PostgreSQL cluster.

Root cause: Connection pool exhaustion due to long-running
transactions from the reporting service.

Mitigation: Killed long-running queries, increased pool size
from 100 to 150 connections, added query timeouts.
```

For **general users**:
```
We are experiencing slower than normal response times for some
features. Our team is working to resolve this quickly.
```

## Warning Signs to Monitor

Watch for these indicators that might trigger using this template:
- P95 latency > 2x normal
- Error rate > 1%
- Connection timeouts
- User reports of slowness
- Database monitoring alerts