惯性聚合 高效追踪和阅读你感兴趣的博客、新闻、科技资讯
阅读原文 在惯性聚合中打开

推荐订阅源

酷 壳 – CoolShell
酷 壳 – CoolShell
aimingoo的专栏
aimingoo的专栏
P
Proofpoint News Feed
宝玉的分享
宝玉的分享
MyScale Blog
MyScale Blog
The GitHub Blog
The GitHub Blog
钛媒体:引领未来商业与生活新知
钛媒体:引领未来商业与生活新知
月光博客
月光博客
量子位
博客园 - 司徒正美
V
V2EX
I
InfoQ
奇客Solidot–传递最新科技情报
奇客Solidot–传递最新科技情报
Vercel News
Vercel News
H
Hackread – Cybersecurity News, Data Breaches, AI and More
美团技术团队
N
Netflix TechBlog - Medium
L
LangChain Blog
IT之家
IT之家
Blog — PlanetScale
Blog — PlanetScale
Cyber Security Advisories - MS-ISAC
Cyber Security Advisories - MS-ISAC
Stack Overflow Blog
Stack Overflow Blog
A
About on SuperTechFans
Microsoft Azure Blog
Microsoft Azure Blog

DEV Community

Authentication Security Deep Dive: From Brute Force to Salted Hashing (With Java Examples) Why AI Systems Don’t Fail — They Drift Spilling beans for how i learn for exam😁"Reinforcement Learning Cheat Sheet" I Replaced Chrome with Safari for AI Browser Automation. Here's What Broke (and What Finally Worked) How Python Borrows Other People's Work The $40 Architecture: Processing 1 Billion API Requests with 99.99% Uptime Vibe Coding: A Workflow Guide (From Zero to SaaS) Most webhook security guides protect the wrong side. The scary part is delivery. Headless CMS for TanStack Start: Build a Blog with Cosmic EU Age Verification App "Hacked in 2 Minutes" — What Actually Happened Comfy Cloud’s delete function does not actually remove files Running AI Models on GPU Cloud Servers: A Beginner Guide Event-driven media intelligence with AWS Step Functions and Bedrock I scored 500 AI prompts across 8 quality dimensions — here's what broke How to Call Google Gemini API from Next.js (Free Tier, No Backend Needed) The Portal Protocol: Reclaiming Human Connection in the Age of AI How to Fix Your Team's Scattered Knowledge Problem With a Self-Hosted Forum Intro to tc Cloud Functors: A Graph-First Mental Model for the Modern Cloud Designing Multi-Tenant Backends With Both Ownership and Team Access I Built a Neumorphic CSS Library with 77+ Components — Here's What I Learned PostgreSQL Performance Optimization: Why Connection Pooling Is Critical at Scale Cómo construí un SaaS multi-rubro para gestionar expensas en Argentina con FastAPI + Vue 3 🚀 I Built an Ethical Hacking Scanner Tool – Open Source Project I Replaced /usage and /context in Claude Code With a Single Statusline A Pythonic Way to Handle Emails (IMAP/SMTP) with Auto-Discovery and AI-Ready Design I Collected 8.9 Million Polymarket Price Points — Here's What I Found About How Markets Really Move EcoTrack AI — Carbon Footprint Tracker & Dashboard Everyone's Using AI. No One Agrees How. 5 self-hosted ebook managers worth trying in 2026 Building Your First AI Agent with LangChain: From Chatbot to Autonomous Assistant
PostgreSQL Row Level Security: The Right Way to Lock Down...
kanta13jp1 · 2026-04-28 · via DEV Community

kanta13jp1

PostgreSQL Row Level Security: The Right Way to Lock Down Your Data

Row Level Security (RLS) enforces access control inside the database, not the application layer. After running 12 parallel AI instances touching the same Supabase database, this is the pattern that keeps the data clean.

Why Application-Layer Checks Aren't Enough

App-layer check:
  request → Edge Function → "Does this user have access?" → SQL
  Problem: EF bug / new instance forgets the check → full table exposed

RLS:
  request → SQL runs → PostgreSQL filters automatically → only visible rows returned
  Problem: none. The filter runs inside the engine

Enter fullscreen mode Exit fullscreen mode

RLS-enabled tables return zero rows to anyone with no policy. Deny-by-default is automatic. You opt in to access explicitly.

Basic Pattern: Users See Their Own Data

-- Enable RLS
ALTER TABLE user_notes ENABLE ROW LEVEL SECURITY;

-- SELECT: own rows only
CREATE POLICY "users_select_own" ON user_notes
  FOR SELECT
  USING (auth.uid() = user_id);

-- INSERT: must match own user_id
CREATE POLICY "users_insert_own" ON user_notes
  FOR INSERT
  WITH CHECK (auth.uid() = user_id);

-- UPDATE: own rows, own user_id only
CREATE POLICY "users_update_own" ON user_notes
  FOR UPDATE
  USING (auth.uid() = user_id)
  WITH CHECK (auth.uid() = user_id);

-- DELETE: own rows only
CREATE POLICY "users_delete_own" ON user_notes
  FOR DELETE
  USING (auth.uid() = user_id);

Enter fullscreen mode Exit fullscreen mode

USING = which rows are visible. WITH CHECK = which rows can be written. UPDATE needs both.

How auth.uid() Works

-- auth.uid() is parsed from the JWT automatically
-- Supabase client sends Authorization: Bearer <token>
-- PostgreSQL resolves auth.uid() from the claim

-- Verify in psql
SELECT auth.uid();   -- returns current session's user_id
SELECT auth.role();  -- 'anon' or 'authenticated'

Enter fullscreen mode Exit fullscreen mode

Edge Functions use the Service Role Key, so auth.uid() returns NULL there. EFs bypass RLS entirely — which means EFs are responsible for their own access checks.

Shared Data Pattern

-- Notes: public OR own
CREATE POLICY "notes_select" ON notes
  FOR SELECT
  USING (
    is_public = true
    OR auth.uid() = user_id
  );

Enter fullscreen mode Exit fullscreen mode

Admin Pattern

CREATE TABLE admin_users (user_id UUID PRIMARY KEY);

CREATE POLICY "admin_select_all" ON user_notes
  FOR SELECT
  USING (
    auth.uid() = user_id
    OR EXISTS (
      SELECT 1 FROM admin_users WHERE user_id = auth.uid()
    )
  );

Enter fullscreen mode Exit fullscreen mode

The EXISTS subquery hits the PK index. Fast even at scale.

Tenant Pattern (Teams / Organizations)

CREATE POLICY "org_members_select" ON org_documents
  FOR SELECT
  USING (
    org_id IN (
      SELECT org_id FROM organization_members
      WHERE user_id = auth.uid()
    )
  );

Enter fullscreen mode Exit fullscreen mode

This project is single-tenant, but if it ever goes multi-tenant SaaS, this is the policy shape.

RLS + Edge Functions: The Right Split

// Service Role Key: bypasses RLS (for admin ops)
const supabaseAdmin = createClient(
  Deno.env.get('SUPABASE_URL')!,
  Deno.env.get('SUPABASE_SERVICE_ROLE_KEY')!,
);

// Pass user JWT: RLS applies automatically
const supabaseUser = createClient(
  Deno.env.get('SUPABASE_URL')!,
  Deno.env.get('SUPABASE_ANON_KEY')!,
  { global: { headers: { Authorization: req.headers.get('Authorization')! } } }
);

// supabaseUser queries are filtered by RLS
// supabaseAdmin queries return everything — handle with care

Enter fullscreen mode Exit fullscreen mode

Debugging RLS

-- List all policies on a table
SELECT schemaname, tablename, policyname, cmd, qual
FROM pg_policies
WHERE tablename = 'user_notes';

-- Test as a specific user
SET LOCAL ROLE authenticated;
SET LOCAL "request.jwt.claims" TO '{"sub": "user-uuid-here"}';
SELECT * FROM user_notes;  -- RLS-filtered result
RESET ROLE;

Enter fullscreen mode Exit fullscreen mode

Performance Notes

1. auth.uid() evaluates per row.

Policies with user_id = auth.uid() use an index — fast. Subquery-heavy policies (e.g. checking team membership) need careful indexing.

2. Cache expensive checks with SECURITY DEFINER functions.

CREATE OR REPLACE FUNCTION is_admin()
RETURNS BOOLEAN
LANGUAGE sql
SECURITY DEFINER
STABLE
AS $$
  SELECT EXISTS (SELECT 1 FROM admin_users WHERE user_id = auth.uid());
$$;

CREATE POLICY "admin_select" ON user_notes
  FOR SELECT
  USING (auth.uid() = user_id OR is_admin());

Enter fullscreen mode Exit fullscreen mode

STABLE lets PostgreSQL cache the result within a single statement.

The Four Rules

  1. ENABLE ROW LEVEL SECURITY on every table. Forgetting this is the #1 RLS mistake.
  2. Trust deny-by-default. Zero policies = zero rows exposed. No fallback needed.
  3. EFs use Service Role Key — they bypass RLS. Write your own checks inside EFs, or switch to the user JWT.
  4. Index user_id. The RLS filter runs on every query; the index makes it free.

Don't put auth logic in the application layer. Put it in the database, where it can't be skipped.