docs: Add PostgreSQL session storage sample and documentation

Merge https://github.com/google/adk-python/pull/3926

### Link to Issue or Description of Change

**1. Link to an existing issue (if applicable):**

- Related: #3916

**2. Or, if no issue exists, describe the change:**

**Problem:**
While `DatabaseSessionService` already supports PostgreSQL through SQLAlchemy, there is no documentation or sample code showing users how to configure and use it.

**Solution:**
Add a comprehensive sample under `contributing/samples/postgres_session_service/` that demonstrates:
- How to configure `DatabaseSessionService` with PostgreSQL
- The auto-generated database schema (sessions, events, app_states, user_states tables)
- Connection URL format and configuration options
- A working sample agent with session persistence

### Testing Plan

**Unit Tests:**

- [x] I have added or updated unit tests for my change.
- [x] All unit tests pass locally.

This is a documentation-only change (new sample), so no new unit tests are required. Existing tests continue to pass.

**Manual End-to-End (E2E) Tests:**

Tested locally with the following steps:

1. Started PostgreSQL using `docker compose up -d`
2. Set environment variables:

```bash
export POSTGRES_URL=postgresql+asyncpg://postgres:postgres@localhost:5432/adk_sessions
export GOOGLE_CLOUD_PROJECT=$(gcloud config get-value project)
export GOOGLE_CLOUD_LOCATION=us-central1
export GOOGLE_GENAI_USE_VERTEXAI=true
```
3. Ran `pip install google-adk asyncpg greenlet` to install the required packages
4. Ran `python main.py` - session created successfully
5. Ran `python main.py` again - previous session resumed with event history
6. Verified tables and rows created in PostgreSQL (sessions, events, app_states, user_states)

### Checklist

- [x] I have read the https://github.com/google/adk-python/blob/main/CONTRIBUTING.md document.
- [x] I have performed a self-review of my own code.
- [x] I have commented my code, particularly in hard-to-understand areas.
- [x] I have added tests that prove my fix is effective or that my feature works.
- [x] New and existing unit tests pass locally with my changes.
- [x] I have manually tested my changes end-to-end.
- [x] Any dependent changes have been merged and published in downstream modules.

### Additional context

This PR adds documentation and a working sample for an already-supported feature. The DatabaseSessionService class already handles PostgreSQL through its DynamicJSON type decorator which uses JSONB for PostgreSQL.

Files added:
- contributing/samples/postgres_session_service/README.md - Comprehensive guide
- contributing/samples/postgres_session_service/agent.py - Sample agent
- contributing/samples/postgres_session_service/main.py - Usage example
- contributing/samples/postgres_session_service/compose.yml - Local PostgreSQL setup
- contributing/samples/postgres_session_service/\_\_init\_\_.py - Package init

Co-authored-by: Liang Wu <wuliang@google.com>
COPYBARA_INTEGRATE_REVIEW=https://github.com/google/adk-python/pull/3926 from hiroakis:feat-support-pg-for-conversation a5279d4fb2a63699f384c670928d77d882c25a25
PiperOrigin-RevId: 868816317
This commit is contained in:
Hiroaki Sano
2026-02-11 13:25:24 -08:00
committed by Copybara-Service
parent 61c329f8ce
commit 657acfadbb
5 changed files with 376 additions and 0 deletions
@@ -0,0 +1,198 @@
# Using PostgreSQL with DatabaseSessionService
This sample demonstrates how to configure `DatabaseSessionService` to use PostgreSQL for persisting sessions, events, and state.
## Overview
ADK's `DatabaseSessionService` supports multiple database backends through SQLAlchemy. This guide shows how to:
- Set up PostgreSQL as the session storage backend
- Configure async connections with `asyncpg`
- Understand the auto-generated schema
- Run the sample agent with persistent sessions
## Prerequisites
- **PostgreSQL Database**: A running PostgreSQL instance (local or cloud)
- **asyncpg**: Async PostgreSQL driver for Python
## Installation
Install the required Python packages:
```bash
pip install google-adk asyncpg greenlet
```
## Database Schema
`DatabaseSessionService` automatically creates the following tables on first use:
### sessions
| Column | Type | Description |
| ----------- | ------------ | ------------------------------ |
| app_name | VARCHAR(128) | Application identifier (PK) |
| user_id | VARCHAR(128) | User identifier (PK) |
| id | VARCHAR(128) | Session UUID (PK) |
| state | JSONB | Session state as JSON |
| create_time | TIMESTAMP | Creation timestamp |
| update_time | TIMESTAMP | Last update timestamp |
### events
| Column | Type | Description |
| ------------------ | ------------ | ------------------------------ |
| id | VARCHAR(256) | Event UUID (PK) |
| app_name | VARCHAR(128) | Application identifier (PK) |
| user_id | VARCHAR(128) | User identifier (PK) |
| session_id | VARCHAR(128) | Session reference (PK, FK) |
| invocation_id | VARCHAR(256) | Invocation identifier |
| timestamp | TIMESTAMP | Event timestamp |
| event_data | JSONB | Event content as JSON |
### app_states
| Column | Type | Description |
| ----------- | ------------ | ------------------------------ |
| app_name | VARCHAR(128) | Application identifier (PK) |
| state | JSONB | Application-level state |
| update_time | TIMESTAMP | Last update timestamp |
### user_states
| Column | Type | Description |
| ----------- | ------------ | ------------------------------ |
| app_name | VARCHAR(128) | Application identifier (PK) |
| user_id | VARCHAR(128) | User identifier (PK) |
| state | JSONB | User-level state |
| update_time | TIMESTAMP | Last update timestamp |
### adk_internal_metadata
| Column | Type | Description |
| ----------- | ------------ | ------------------------------ |
| key | VARCHAR(128) | Metadata key |
| value | VARCHAR(256) | Metadata value |
## Configuration
### Connection URL Format
```python
postgresql+asyncpg://username:password@host:port/database
```
### Basic Usage
```python
from google.adk.sessions.database_session_service import DatabaseSessionService
from google.adk.runners import Runner
# Initialize with PostgreSQL URL
session_service = DatabaseSessionService(
"postgresql+asyncpg://postgres:postgres@localhost:5432/adk_sessions"
)
# Use with Runner
runner = Runner(
app_name="my_app",
agent=my_agent,
session_service=session_service,
)
```
### Advanced Configuration
Pass additional SQLAlchemy engine options:
```python
session_service = DatabaseSessionService(
"postgresql+asyncpg://postgres:postgres@localhost:5432/adk_sessions",
pool_size=10,
max_overflow=20,
pool_timeout=30,
pool_recycle=1800,
)
```
## Running the Sample
### 1. Start PostgreSQL
Using Docker:
```bash
docker compose up -d
```
Or use an existing PostgreSQL instance.
### 2. Configure Connection
Create a `.env` file:
```bash
POSTGRES_URL=postgresql+asyncpg://postgres:postgres@localhost:5432/adk_sessions
GOOGLE_CLOUD_PROJECT=<your-gcp-project-id>
GOOGLE_CLOUD_LOCATION=us-central1
GOOGLE_GENAI_USE_VERTEXAI=true
```
Or run export command.
```bash
export POSTGRES_URL=postgresql+asyncpg://postgres:postgres@localhost:5432/adk_sessions
export GOOGLE_CLOUD_PROJECT=$(gcloud config get-value project)
export GOOGLE_CLOUD_LOCATION=us-central1
export GOOGLE_GENAI_USE_VERTEXAI=true
```
### 3. Run the Agent
```bash
python main.py
```
Or use the ADK:
```bash
adk run .
```
## Session Persistence
Sessions and events are persisted across application restarts:
```python
# First run - creates a new session
session = await session_service.create_session(
app_name="my_app",
user_id="user1",
session_id="persistent-session-123",
)
# Later run - retrieves the existing session
session = await session_service.get_session(
app_name="my_app",
user_id="user1",
session_id="persistent-session-123",
)
```
## State Management
PostgreSQL's JSONB type provides efficient storage for state data:
- **Session state**: Stored in `sessions.state`
- **User state**: Stored in `user_states.state`
- **App state**: Stored in `app_states.state`
## Production Considerations
1. **Connection Pooling**: Use `pool_size` and `max_overflow` for high-traffic applications
2. **SSL/TLS**: Always use encrypted connections in production
3. **Backups**: Implement regular backup strategies for session data
4. **Indexing**: The default schema includes primary key indexes; add additional indexes based on query patterns
5. **Monitoring**: Monitor connection pool usage and query performance
@@ -0,0 +1,16 @@
# Copyright 2026 Google LLC
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
from . import agent
@@ -0,0 +1,43 @@
# Copyright 2026 Google LLC
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
"""Sample agent demonstrating PostgreSQL session persistence."""
from datetime import datetime
from datetime import timezone
from google.adk.agents.llm_agent import Agent
def get_current_time() -> str:
"""Get the current time.
Returns:
A string with the current time in ISO 8601 format.
"""
return datetime.now(timezone.utc).isoformat()
root_agent = Agent(
model="gemini-2.0-flash",
name="postgres_session_agent",
description="A sample agent demonstrating PostgreSQL session persistence.",
instruction="""
You are a helpful assistant that demonstrates session persistence.
You can remember previous conversations within the same session.
Use the get_current_time tool when asked about the time.
When the user asks what you remember, summarize the previous conversation.
""",
tools=[get_current_time],
)
@@ -0,0 +1,24 @@
# Docker Compose configuration for the postgres_session_service sample.
#
# This file defines a PostgreSQL service used to demonstrate ADK's
# DatabaseSessionService with a persistent backend. It sets up a
# postgres:16-alpine container with:
# - Default credentials (user: postgres, password: postgres)
# - A pre-created database named 'adk_sessions'
# - Port 5432 exposed for local access
# - A named volume 'postgres_data' for data persistence
services:
postgres:
image: postgres:16-alpine
environment:
POSTGRES_USER: postgres
POSTGRES_PASSWORD: postgres
POSTGRES_DB: adk_sessions
ports:
- "5432:5432"
volumes:
- postgres_data:/var/lib/postgresql/data
volumes:
postgres_data:
@@ -0,0 +1,95 @@
# Copyright 2026 Google LLC
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
"""Example demonstrating PostgreSQL session persistence with DatabaseSessionService."""
import asyncio
import os
import agent
from dotenv import load_dotenv
from google.adk.runners import Runner
from google.adk.sessions.database_session_service import DatabaseSessionService
from google.adk.sessions.session import Session
from google.genai import types
load_dotenv(override=True)
async def main():
"""Main function demonstrating PostgreSQL session persistence."""
postgres_url = os.environ.get("POSTGRES_URL")
if not postgres_url:
raise ValueError(
"POSTGRES_URL environment variable not set. "
"Please create a .env file with"
" POSTGRES_URL=postgresql+asyncpg://user:password@localhost:5432/adk_sessions"
)
app_name = "postgres_session_demo"
user_id = "demo_user"
session_id = "persistent-session"
# Initialize PostgreSQL-backed session service
session_service = DatabaseSessionService(postgres_url)
runner = Runner(
app_name=app_name,
agent=agent.root_agent,
session_service=session_service,
)
# Try to get existing session or create new one
session = await session_service.get_session(
app_name=app_name,
user_id=user_id,
session_id=session_id,
)
if session:
print(f"Resuming existing session: {session.id}")
print(f"Previous events count: {len(session.events)}")
else:
session = await session_service.create_session(
app_name=app_name,
user_id=user_id,
session_id=session_id,
)
print(f"Created new session: {session.id}")
async def run_prompt(session: Session, new_message: str):
"""Send a prompt to the agent and print the response."""
content = types.Content(
role="user", parts=[types.Part.from_text(text=new_message)]
)
print(f"User: {new_message}")
async for event in runner.run_async(
user_id=user_id,
session_id=session.id,
new_message=content,
):
if event.content and event.content.parts and event.content.parts[0].text:
print(f"{event.author}: {event.content.parts[0].text}")
print("------------------------------------")
await run_prompt(session, "What time is it? Please remember this.")
print("------------------------------------")
await run_prompt(session, "What did I just ask you?")
print("------------------------------------")
print("\nSession persisted to PostgreSQL. Run again to see event history.")
if __name__ == "__main__":
asyncio.run(main())