Storage Backends

Pluggable storage for FSM state persistence. Choose between in-memory for development, Redis for production, or implement your own.

Choosing a Backend

Backend	Persistence	Distributed	Best For
`MemoryStorage`	❌ In-process only	❌ Single instance	Development, testing, small bots
`RedisStorage`	✅ Survives restarts	✅ Multi-instance	Production, distributed bots
Custom	You decide	You decide	Specialized requirements

BaseStorage Interface

All storage backends implement this abstract interface:

Method	Description
`set_state(id, state, ttl=None)`	Store state data with optional TTL
`get_state(id)`	Retrieve state data by identifier
`delete_state(id)`	Remove state data
`compare_and_set(id, new, expected, ttl)`	Atomic compare-and-swap for concurrency safety
`list_keys(pattern)`	List all keys matching a glob pattern
`clear_namespace()`	Remove all keys in the storage namespace
`start() / stop()`	Lifecycle hooks for connection management
`health()`	Health check returning `True` if storage is operational

Quick Setup

Memory (development)

from pyrogram_patch.fsm import MemoryStorage

manager = await patch(app)
manager.set_storage(MemoryStorage(default_ttl=3600))  # 1h TTL

Redis (production)

import redis.asyncio as redis
from pyrogram_patch.fsm import RedisStorage

redis_client = redis.Redis(host="localhost", port=6379)
manager = await patch(app)
manager.set_storage(RedisStorage(redis_client, prefix="mybot:"))