Introducing ReSight
The Redis Performance Dashboard
Your Team Has Been Waiting For
Stop flying blind on your cache. ReSight gives you real-time Redis visibility, historical trend analysis, intelligent Slack alerting, and collaborative team tooling — all in a single, production-safe dashboard.
Redis is the heartbeat of modern applications — session storage, caching, queues, rate limiting. When it slows down or misbehaves, everything suffers. And diagnosing Redis performance issues without the right tooling means hunting through redis-cli output, comparing notes across SSH sessions, and trying to remember what the cache hit ratio was an hour ago. That's no way to run production.
ReSight was built to change that. It's a self-hosted Redis performance profiler that connects directly to your Redis instances, collects over 40 metrics in real time, stores time-series history in PostgreSQL, fires Slack alerts when thresholds are breached, and generates polished PDF reports — all from a clean, accessible web interface. Whether you're profiling a single server during a load test or watching a fleet of six instances across staging and production, ReSight has you covered.
Deep Redis Visibility
40+ metrics across memory, latency, operations, network, persistence, CPU, and clients — organized into focused panels that tell a story.
Built for Teams
Shared Redis connections, role-based access control, Slack integration, and invite links — without handing out credentials.
Production Safe
Read-only monitoring. No writes to Redis. No config changes required. AES-256-GCM credential encryption. Tiered rate limiting on every endpoint.
Instant Setup
One docker compose up command. No agents, no SDKs, no instrumentation changes. Schema initializes automatically.
Know Your Redis Health at a Glance
Every ReSight dashboard opens with a Health Score from 0 to 100 — a single composite number that tells you immediately whether your Redis server needs attention. The score weighs four factors that matter most:
- Cache Hit Ratio: Are requests being served from Redis or falling through to your database?
- P50 Latency: How fast is Redis actually responding?
- CPU Utilization: Is Redis under computational pressure?
- Throughput: Are operations per second trending in the right direction?
Color-coded green, amber, and red, the health score gives you an instant gut-check on whether your Redis instance is healthy or needs investigation — without having to read every panel every time.
Always-Visible KPI Strip
No matter what you're viewing, a compact KPI bar at the top of every dashboard always shows your four most critical signals: Cache Hit Ratio, Total Key Count, P50 Latency, and Uptime. The hit ratio is color-coded green/amber/red at a glance — and clicking the Uptime chip opens a full Redis restart history timeline.
Real-Time Monitoring That Doesn't Burden Your Redis
ReSight polls your Redis server every 10 seconds in normal mode. Flip on Real-Time Mode during load tests or incidents and it drops to 5-second intervals — giving you near-live visibility when you need it most.
Each monitoring session is independent, so different team members can run at different cadences without interference. A shared TCP connection cache means multiple engineers monitoring the same Redis server produce no additional load multiplication — connections are reused across the team.
Session state persists in browser localStorage — your monitoring sessions survive page reloads and tab restores, so you don't lose your place during an incident.
A Dashboard Built for Real-World Debugging
The moment you connect to a Redis server, ReSight starts collecting metrics. Panels are organized around the questions engineers actually ask when something feels slow:
🧠 Memory Usage & Analysis
Used, peak, and total memory with utilization percentage. Fragmentation ratio, wasted memory, evicted and expired key counts — the full picture of where memory is going.
⚡ Operations Breakdown
Read, write, and delete distribution as a pie chart. Instantly understand whether your workload is cache-heavy or write-heavy and how that's shifting.
🎯 Keyspace & Hit Ratio
Cache hits vs. misses visualized and tracked over time. A low hit ratio is often the first sign that TTLs are wrong or keys are being evicted prematurely.
📊 Latency Percentiles
P50, P95, and P99 — not just the average. Outliers hide in averages. Percentile tracking surfaces the slow tail that's actually hurting your SLA.
🐢 Slow Command Log
The last 10 slow commands pulled directly from Redis's SLOWLOG, paginated and easy to scan. No more redis-cli slowlog get in a terminal session.
🔑 Low Hit Ratio Keys
ReSight scans your keyspace and surfaces individual keys with an estimated hit ratio below 75%. Find exactly which keys are underperforming — not just that the overall ratio is low.
🌐 Network I/O
Real-time input/output throughput in KB/s, total bytes transferred, and I/O ratio. Essential for diagnosing bandwidth bottlenecks and oversized value transfers.
💾 Persistence Metrics
RDB and AOF status, last save times, and fork metrics. Know whether your persistence strategy is working and how much it's impacting performance.
🔌 Client Connections
Connected, blocked, and rejected client counts with utilization status. Blocked and rejected clients flag connection pool exhaustion or misconfigured timeouts.
🖥️ CPU Metrics
System and user CPU time with an efficiency rating. CPU spikes in Redis often signal expensive commands or misconfigured key scans — now you'll see them the moment they happen.
Every metric comes with color-coded severity indicators. Every panel includes trend context. ReSight doesn't just show you numbers — it tells you whether those numbers are good, concerning, or critical.
Monitor Your Entire Redis Fleet
Most teams don't have just one Redis server. You might have a cache instance, a session store, a queue backend, and separate instances for staging and production. ReSight lets you monitor up to six Redis instances simultaneously.
The multi-instance overview shows a compact metric card for each server in a grid layout — health score, hit ratio, latency, and key count at a glance. Click any card to expand it into a full detailed dashboard for that instance. Staggered polling across sessions prevents request bursts, and the shared connection cache means team members monitoring the same instance don't multiply the Redis load.
Alert rules evaluate across all active sessions simultaneously, with per-rule per-session cooldowns ensuring you get notified without being flooded.
Historical Trends: See What Happened, Not Just What's Happening
Real-time snapshots show you now. History explains why. ReSight stores time-series metric snapshots to PostgreSQL every 60 seconds and retains them for up to 30 days (configurable).
Historical Metrics lets you chart key signals over selectable ranges — 1H, 6H, 24H, 7D, and 30D — with automatic downsampling for longer ranges so charts stay fast and readable no matter how much data you have. Trend indicators and historical comparisons surface whether a metric is improving or degrading over time.
This is the feature that separates a monitoring tool from a debugging tool. When your on-call engineer asks "was this already slow before the deploy?" — now you have the answer.
Uptime & Reboot Tracking
Redis restarts are a silent source of performance problems — a surprise reboot clears your entire cache, causing a hit ratio crash and a thundering herd of cache misses hammering your database. ReSight automatically detects restart events by comparing current uptime against the previous reading.
Every restart is logged with a timestamp and the uptime value before the event. Click the Uptime KPI to open a full restart history timeline. Correlate any hit ratio drop with a reboot in seconds.
Alert Rules & Slack Integration: Your 24/7 On-Call Teammate
ReSight's Slack integration covers two distinct use cases: proactive threshold alerting and ad-hoc panel sharing.
Cache hit ratio, response time, CPU, memory usage, fragmentation ratio, ops/sec, client connections (connected/blocked/rejected), network I/O, evicted keys, and memory utilization. Set any threshold with >, <, ≥, or ≤ operators.
Configure a cooldown period per rule so you're not flooded during sustained incidents. Severity is auto-calculated and each rule targets a specific Slack channel.
Every alert firing is recorded — what metric, what value, when it triggered, which Slack thread received it. You'll always know what fired and why.
Every dashboard panel has a hover-reveal share button. Send data to any channel or thread with an optional comment. Charts are shared as images, lists are formatted with backtick-wrapped values. The fastest way to pull a colleague into a debugging session.
Up to 50 alert rules per user. The standalone Alert Rules page gives a clean management interface for viewing, editing, and toggling rules across all your Redis connections.
PDF Performance Reports: From Engineers to Executives
Need to document a performance review or share findings with stakeholders? ReSight generates professional PDF reports entirely in the browser — no server-side rendering, no data leaving your infrastructure. Two report variants address different audiences:
Internal Engineering Report
Full technical analysis including server configuration, network I/O and persistence details, top commands by frequency, per-database keyspace breakdown, color-coded severity indicators, and prioritized action items (HIGH / MEDIUM / LOW). Built for engineering teams doing post-incident reviews or capacity planning.
Customer / Stakeholder Report
A sanitized version omitting operational internals. Plain-language health assessment, key findings in accessible language, and remediation steps without jargon. Safe to hand directly to clients or non-technical management without revealing infrastructure details.
Both reports include a prioritized action plan with concrete remediation steps — not just "cache hit ratio is low" but exactly what to investigate and what to change.
Secure Connection Management
Connection to a Redis server is as simple as entering a connection string. ReSight supports password-protected instances and lets you save connections with custom names for one-click reconnection. Security is built into the connection layer from the start:
AES-256-GCM Encryption
Redis passwords are encrypted before storage in PostgreSQL. Decryption happens server-side only — credentials are never exposed to the browser after the initial connection.
User-Scoped Data Isolation
Connections, metrics, and alert rules are strictly isolated per user. You'll only ever see your own saved connections and their history — never another user's data.
Connection History & Reconnect
A collapsible history panel shows recent connections with metadata. Team-shared connections appear in a separate section. One-click reconnection from saved credentials.
Team Collaboration Without Shared Credentials
Team Management
Create teams and share specific Redis connections with specific groups. Role-based access: Admins can invite, remove, and share; Viewers can monitor and view metrics. Single-use invite links with 7-day expiry and optional email domain restriction.
Zero Credential Exposure
Team members can monitor shared Redis servers from their own accounts without ever seeing the underlying credentials. The connection owner retains full control and can revoke team access at any time.
Google SSO
Authentication via Google OAuth 2.0 with configurable allowed email domains. JWT sessions with 24-hour expiry. Per-user rate limiting on all API endpoints — monitor (120/min), general (100/min), auth (10/min), Slack (10/min).
nginx Security Headers
X-Frame-Options, X-Content-Type-Options, X-XSS-Protection, Referrer-Policy, and Permissions-Policy configured out of the box. CORS scoped to your configured origin. Non-root nginx container.
Designed for Everyone: Dark Mode & Full Accessibility
ReSight is built on React 18 with Tailwind CSS and Radix UI primitives — a modern, professional interface that works in any context and for any user:
- Dark and Light mode — system-aware by default, with a manual toggle that persists across sessions
- Collapsible sidebar — icons-only when collapsed, full labels when expanded; state remembered across reloads
- Semantic HTML landmarks and ARIA labels on all interactive elements and charts for screen reader compatibility
- Skip-to-content link for keyboard users
- Screen reader text alternatives for all color-coded severity indicators
- Full keyboard navigation with visible focus-outline indicators
- In-app User Guide — comprehensive docs accessible directly from the sidebar, with a table of contents and scroll-to-section navigation
Every panel is wrapped in React error boundaries so a single component failure never takes down the entire dashboard — critical during high-stakes incident debugging.
Everything ReSight Monitors
Who ReSight Is For
Backend Engineers
Profile Redis performance during development and load testing. Find slow commands, identify low hit ratio keys, and validate TTL strategies before shipping to production.
Platform & SRE Teams
Monitor production Redis fleets with historical trend analysis, threshold alerts to Slack, and reboot detection. Know about problems before your users do.
DevOps & Ops Engineers
Get Redis observability without adding a full APM suite. Self-hosted, Docker-native, and configurable without touching application code or Redis configuration.
Consultants & Agencies
Generate professional PDF performance reports for clients. Full technical report for your team, plain-language summary for stakeholders — both from the same dashboard session.
Up and Running in Minutes
ReSight runs as a 3-container Docker stack — an nginx frontend, an Express API server, and a PostgreSQL 16 database for time-series metric storage. No agents. No sidecars. No Redis configuration changes. Bring it up with three commands:
cp .env.example .env
# Add your Google OAuth credentials and encryption key
docker compose up -d --build
# Visit http://localhost:3001
The database schema initializes automatically on first startup. Metrics are retained for 30 days by default — configurable via environment variable. Key scanning can be disabled entirely for high-throughput production clusters.
Your Redis deserves better visibility.
ReSight brings production-grade Redis observability to teams of any size — without the cloud lock-in, the vendor data sharing, or the per-metric pricing. Self-host it, own your data, and finally know what your cache is doing.