Horizontal Sharding for High-Load PostgreSQL Web Applications

Note: when the number of records in the orders table exceeds 200 million and write load reaches 5000 transactions per second, PostgreSQL on a single server can't handle it: latency grows to 100 ms, checkpoints slow to several minutes, disk is full (10 TB). You've already tried date-based partitioning, master-slave replication, and Redis caching — but write conflicts and lock contention remain. The only option left is horizontal database sharding. We design and implement such solutions for high-load web applications. Our experience: over 50 projects with distributed systems, 8 years of practice, and we guarantee reliability. Order a database audit for $499 — we'll find the bottlenecks and suggest the optimal architecture, potentially saving you up to 40% on infrastructure costs. Typical savings: $5,000/month on infrastructure.

Partitioning vs Sharding: Which to Choose?

Partitioning splits a single table into physical parts within one PostgreSQL instance. Sharding distributes data across multiple independent servers. Partitioning is 10x simpler than sharding and is often sufficient — start with it. According to PostgreSQL Documentation, partitioning is recommended for tables over 100 GB.

-- Range partitioning by date (logs, events)
CREATE TABLE events (
    id         BIGSERIAL,
    user_id    BIGINT       NOT NULL,
    event_type VARCHAR(50)  NOT NULL,
    created_at TIMESTAMPTZ  NOT NULL,
    data       JSONB
) PARTITION BY RANGE (created_at);

CREATE TABLE events_2024_q1 PARTITION OF events
    FOR VALUES FROM ('2024-01-01') TO ('2024-04-01');

CREATE TABLE events_2024_q2 PARTITION OF events
    FOR VALUES FROM ('2024-04-01') TO ('2024-07-01');

-- Hash partitioning for even distribution
CREATE TABLE user_sessions (
    id      BIGSERIAL,
    user_id BIGINT NOT NULL,
    token   VARCHAR(255) NOT NULL,
    data    JSONB
) PARTITION BY HASH (user_id);

CREATE TABLE user_sessions_0 PARTITION OF user_sessions
    FOR VALUES WITH (MODULUS 4, REMAINDER 0);
-- etc. up to REMAINDER 3

If partitioning no longer helps (write load hits CPU, data doesn't fit on disk), move to sharding.

How to Choose the Shard Key?

The shard key is the main architectural decision. Good options: user_id for user-centric apps, tenant_id for multi-tenant SaaS, region for geographically distributed data. Bad options: created_at — hot spot on the last shard, status — uneven distribution, UUID v4 — no locality, poor cache hit.

Why Use Citus Instead of Custom Sharding?

Citus is a PostgreSQL extension that turns it into a distributed DB. Citus reduces implementation time by 5x compared to custom sharding because it automatically manages distribution, rebalancing, and JOIN locality. The Citus Enterprise license costs ~$1,000 per month, but infrastructure savings can be $5,000 per month due to reduced server count by 30%.

-- Add workers
SELECT citus_add_node('worker1', 5432);
SELECT citus_add_node('worker2', 5432);

-- Create distributed table
CREATE TABLE orders (
    id         BIGSERIAL,
    tenant_id  INT          NOT NULL,
    user_id    BIGINT       NOT NULL,
    status     VARCHAR(20)  NOT NULL,
    total      DECIMAL(12,2),
    created_at TIMESTAMPTZ  NOT NULL DEFAULT NOW(),
    PRIMARY KEY (id, tenant_id)
);

SELECT create_distributed_table('orders', 'tenant_id', shard_count => 32);

-- Colocated table (JOIN by tenant_id is local)
CREATE TABLE order_items (
    id         BIGSERIAL,
    tenant_id  INT    NOT NULL,
    order_id   BIGINT NOT NULL,
    product_id BIGINT NOT NULL,
    quantity   INT    NOT NULL,
    PRIMARY KEY (id, tenant_id)
);

SELECT create_distributed_table('order_items', 'tenant_id', colocate_with => 'orders');

-- Reference table: replicated to all workers
CREATE TABLE categories (id BIGSERIAL PRIMARY KEY, name VARCHAR(200));
SELECT create_reference_table('categories');

After this, queries with a filter on tenant_id are routed to the specific shard. JOINs between orders and order_items by tenant_id execute locally on the worker.

Comparison of approaches:

Custom Sharding: When Full Control?

Without Citus (or when complete control is needed), we implement sharding at the application level. Use consistent hashing with 150 virtual nodes — this minimizes data movement during resharding to only 1/N of data.

# sharding/router.py
import hashlib
from dataclasses import dataclass
from typing import Any

@dataclass
class ShardConfig:
    host: str
    port: int
    database: str

SHARDS: dict[int, ShardConfig] = {
    0: ShardConfig('db-shard-0', 5432, 'myapp_0'),
    1: ShardConfig('db-shard-1', 5432, 'myapp_1'),
    2: ShardConfig('db-shard-2', 5432, 'myapp_2'),
    3: ShardConfig('db-shard-3', 5432, 'myapp_3'),
}
SHARD_COUNT = len(SHARDS)

def get_shard_id(shard_key: Any) -> int:
    key_bytes = str(shard_key).encode('utf-8')
    hash_value = int(hashlib.md5(key_bytes).hexdigest(), 16)
    return hash_value % SHARD_COUNT

def get_shard_config(shard_key: Any) -> ShardConfig:
    return SHARDS[get_shard_id(shard_key)]

Connections to shards:

from contextlib import contextmanager
from sqlalchemy import create_engine
from sqlalchemy.orm import sessionmaker
from functools import lru_cache

@lru_cache(maxsize=None)
def _get_engine(shard_id: int):
    cfg = SHARDS[shard_id]
    dsn = f"postgresql+psycopg2://user:pass@{cfg.host}:{cfg.port}/{cfg.database}"
    return create_engine(dsn, pool_size=5, max_overflow=10)

@contextmanager
def get_shard_session(shard_key):
    shard_id = get_shard_id(shard_key)
    Session = sessionmaker(bind=_get_engine(shard_id))
    session = Session()
    try:
        yield session
        session.commit()
    except Exception:
        session.rollback()
        raise
    finally:
        session.close()

How to Handle Queries Without a Shard Key?

Queries without a shard key are the hardest. Two approaches exist. Scatter-gather queries all shards in parallel: simple implementation, but latency grows linearly with each new shard. Global index stores the mapping in a separate DB: fast lookup, but 2x higher write overhead. Scatter-gather suits rare analytical queries (e.g., once per hour); global index is better if cross-shard queries occur more than 10% of the time.

import asyncio
import asyncpg

async def get_all_orders_by_status(status: str) -> list[dict]:
    async def query_shard(shard_id: int) -> list[dict]:
        cfg = SHARDS[shard_id]
        conn = await asyncpg.connect(host=cfg.host, database=cfg.database, user='app', password='pass')
        rows = await conn.fetch("SELECT * FROM orders WHERE status = $1 ORDER BY created_at DESC LIMIT 100", status)
        await conn.close()
        return [dict(r) for r in rows]
    results = await asyncio.gather(*[query_shard(i) for i in range(SHARD_COUNT)])
    all_orders = [o for shard_result in results for o in shard_result]
    all_orders.sort(key=lambda x: x['created_at'], reverse=True)
    return all_orders[:100]

Process of Work

1. Analyze current load and bottlenecks: measure write throughput, latency, database size, query patterns.
2. Design schema: choose shard key, number of shards, replication strategy.
3. Develop router and migrate data: implement routing (Citus or application-level), move data with minimal downtime.
4. Load testing: simulate peak load, verify latency and throughput.
5. Deploy and monitor: set up alerts for hot spots, slow queries, rebalancing failures.

Resharding: How to Add a New Shard Without Downtime?

With consistent hashing and virtual nodes (vnodes), only ~1/N data moves (e.g., 25% when adding a 4th shard to 3). Citus automatically redistributes data via citus_rebalance_start(). Without Citus, the process is more complex: stop the application (allow 30 min downtime), redistribute data across the new ring, update the router configuration. To minimize downtime, use a gradual migration with read-only old shards (downtime < 5 min).

What's Included

• Architectural diagram of the distributed DB with shard keys and routing scheme.
• Shard configuration (PostgreSQL settings, connection pools, monitoring).
• Router implementation at application level or via Citus.
• Monitoring setup (Prometheus + Grafana) for hot spots and latency.
• Operations and recovery documentation.
• Team training on distributed schema.
• 30-day post-launch support.

coordinator: 4 vCPU, 16 GB RAM, SSD
worker1: 8 vCPU, 32 GB RAM, NVMe
worker2: 8 vCPU, 32 GB RAM, NVMe
shard_count: 64
replication_factor: 2

Estimated Timelines and Pricing

Pricing is calculated individually. Infrastructure savings from proper sharding can reach 40% — e.g., reducing AWS bill from $10k to $6k per month. Get a consultation (free 30 min) — we'll analyze your load and propose the optimal architecture. We provide turnkey sharding solutions in 3-5 days. Contact us for a free project estimate.

We also recommend reading about Consistent hashing and Citus documentation.

Backend Development Services: Laravel, Node.js, Go, Django, PostgreSQL

On a production server at 3:14 AM, the Laravel Jobs queue stopped processing. 40,000 unprocessed jobs in Redis. Cause: worker crashed due to a memory leak in one of the Jobs (leak via a static variable in an Eloquent observer), supervisor didn't restart it because of misconfigured stopwaitsecs. This is not a hypothetical scenario — it's Tuesday. We analyzed such an incident on a project with 500 RPS load: diagnosis took 4 hours, fix — 20 minutes. So you don't lose money on downtime, we offer backend development services with a focus on production-grade reliability. We'll assess your project in 2 days.

Backend is what works when no one is watching. Or doesn't work. We guarantee you'll have the first option.

How do we ensure production-grade reliability from day one?

What we do correctly from day one

Service Layer over Fat Controllers. Controller receives HTTP request, validates it via Form Request, passes data to Service, returns response. Business logic in Service, not Controller. This sounds trivial, but most legacy projects have controllers with 500 lines and SQL queries inside.

Repository Pattern we use cautiously. If you just wrap Model::where(...) in a repository method — that's boilerplate without benefit. Repository is justified when: you need to abstract from the data source (DB + cache + external API) or when query logic is complex enough to isolate.

Jobs, Events, Listeners. Everything that can be async — make async. Sending email, PDF generation, external API sync, aggregate recalculation — into Queue. Laravel Horizon for queue monitoring in Redis: see throughput, failed jobs, processing time per queue.

How Octane handles high load

Laravel Octane with RoadRunner or Swoole keeps the app in memory between requests — removes bootstrap overhead (config loading, class autoloading) on each HTTP request. Gain: 3–8x on synthetic benchmarks, 2–4x on real applications. Important: no state between requests in static variables — that leads to exactly the incidents from the beginning. We use this in projects with >1000 RPS.

What to do about N+1 queries

N+1 is the most common cause of slow pages in Laravel apps. Standard story: page worked fine on dev with 10 records, on production with 10,000 — 8-second load.

Laravel Debugbar in dev environment shows the number of queries per page. More than 20 queries per page — signal for audit.

Model::preventLazyLoading(! app()->isProduction());

Telescope for profiling in staging: logs all queries, jobs, mail, notifications with time detail. Numbers: after implementing eager loading, page load time drops from 8s to 0.3s — 27 times faster.

PostgreSQL: indexes that are actually needed

PostgreSQL 14+ is the primary DB on all projects. We use PgBouncer + PostgreSQL combination. 10+ years experience, more than 50 backend projects, 5 years on the market.

How PostgreSQL helps avoid slow queries

Composite indexes for frequent WHERE + ORDER BY. If you have WHERE user_id = ? AND status = ? ORDER BY created_at DESC — you need (user_id, status, created_at DESC). A separate index on (user_id) doesn't help much with sorting.

Partial indexes. If 95% of queries go with WHERE status = 'active':

CREATE INDEX idx_orders_active ON orders (created_at DESC)
WHERE status = 'active';

The index is small, fast, covers the main load.

GIN indexes for JSONB and arrays. @> operator without GIN index — seq scan. With index — fast even on millions of rows.

GIN for full-text search. to_tsvector + GIN instead of LIKE '%query%'. LIKE without index is always seq scan. With pg_trgm extension and gin_trgm_ops — supports LIKE with index, useful for CRM search by partial match.

Connection pooling: why it's more important than it seems

Rails, Laravel, Django open a new connection to PostgreSQL for each PHP/Python process. With 100 workers — 100 connections. PostgreSQL starts degrading from 200–300 active connections — overhead on connection management becomes significant.

PgBouncer — connection pooler in front of PostgreSQL. Transaction pooling mode: connection to PostgreSQL is occupied only during a transaction, returned to pool between requests. 1000 application workers → 20–50 actual connections to PostgreSQL. This reduces latency by 40% and hosting costs by 30%.

Node.js with Fastify: when it's better than Laravel

Node.js is justified for:

• Realtime: WebSocket servers, Server-Sent Events, chat, live updates
• Streaming: large files, video, streaming data
• High I/O concurrency: many parallel requests to external APIs without heavy business logic
• Serverless: Lambda/Cloud Functions — Node.js starts faster than PHP

Fastify over Express: 2–3 times faster on benchmarks, built-in JSON Schema validation, better TypeScript support, plugin architecture.

Typical realtime architecture: Laravel — core business logic and REST API. Node.js + Socket.io or ws — WebSocket server. Laravel publishes events to Redis Pub/Sub, Node.js subscribes and broadcasts to clients. This separation allows scaling the WebSocket server independently of the main app.

Go: microservices and high load

Go we use for:

• High-load microservices (>10,000 RPS)
• Background workers with strict latency requirements
• DevOps tools and CLI
• gRPC services in microservice architecture

Goroutines — thousands of times cheaper than OS threads. 10,000 concurrent connections on Go is normal on one server.

But Go is not a silver bullet. Development is slower than Laravel: more boilerplate, no ORM at Eloquent level, error handling with if err != nil everywhere. Justified only when performance is a real requirement, not an assumption.

Django and Python backend

Django with DRF (Django REST Framework) — for tasks where Python is needed: ML pipelines, data processing, integrations with AI tools.

Celery for background tasks — similar to Laravel Queue but more complex to configure. Celery Beat for cron tasks.

Django ORM vs raw SQL: ORM is convenient for CRUD. For analytical queries with multiple JOINs, window functions, and CTEs — connection.execute() with raw SQL is more readable and predictable.

Redis: not just cache

Redis in our projects plays multiple roles:

Redis Cluster for horizontal scaling. Sentinel for automatic failover on standalone setups.

Deployment and infrastructure

Docker + docker-compose — standard for local development and production. Each service in a container: PHP-FPM/Octane, Nginx, PostgreSQL, Redis, Queue Worker, Scheduler.

CI/CD via GitHub Actions:

1. Run tests (PHPUnit / Pest, Vitest, Playwright)
2. Build Docker image
3. Push to Container Registry
4. Deploy: docker pull → docker-compose up -d on server, or Kubernetes rolling update

Zero-downtime deploy for Laravel: php artisan down --secret=TOKEN is not needed with proper configuration. Strategy: new container starts next to the old one, Nginx switches traffic after health check, old container stops.

Monitoring: Sentry for exception tracking with alerting in Slack/Telegram. Grafana + Prometheus (or Grafana Cloud) for metrics: CPU, memory, request rate, queue depth, database connection count. Alerts on: error rate > 1%, p99 latency > 2s, queue depth > 1000 jobs.

What's included in turnkey work

• Architecture design (API documentation, DB schema, service diagram)
• Implementation according to agreed specification with code review
• CI/CD, monitoring, alerting setup
• Load testing (k6, wrk) with report
• Handover of source code, access, deployment instructions
• Training of customer's team (2-3 sessions)
• Warranty support for 1 month after delivery

Timeline benchmarks

Pricing is calculated individually after analyzing load, integrations, and business logic. Contact us for a free audit of your current backend — get an optimization plan in 2 days. Request a consultation.