> ## Documentation Index
> Fetch the complete documentation index at: https://docs.plumi.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# ClickHouse

> Connect to ClickHouse for high-performance analytics

## Overview

Connect Plumi to ClickHouse databases for blazing-fast analytics on large datasets. Supports:

* Self-hosted ClickHouse
* ClickHouse Cloud
* Altinity Cloud

## Connection Details

| Field    | Description                         | Example                  |
| -------- | ----------------------------------- | ------------------------ |
| Host     | ClickHouse server address           | `clickhouse.example.com` |
| Port     | HTTP port (default: 8443 for HTTPS) | `8443`                   |
| Database | Database name                       | `analytics`              |
| Username | Database user                       | `plumi_readonly`         |
| Password | User password                       | `********`               |
| SSL      | Enable SSL (recommended)            | `true`                   |

## Setting Up a Read-Only User

Create a dedicated read-only user:

```sql theme={null}
-- Create user
CREATE USER plumi_readonly IDENTIFIED BY 'your_secure_password';

-- Grant read access
GRANT SELECT ON analytics.* TO plumi_readonly;
```

## ClickHouse Cloud

For ClickHouse Cloud:

1. Go to your ClickHouse Cloud console
2. Navigate to **Settings > Security**
3. Add Plumi's IP to the allowed list
4. Use the provided hostname and credentials

## Best Practices

### Query Optimization

ClickHouse is optimized for analytical queries. For best performance:

* Use aggregations (`GROUP BY`, `SUM`, `COUNT`)
* Filter by partitioning columns first
* Avoid `SELECT *` on large tables
* Use `LIMIT` when exploring data

### Example Queries

```sql theme={null}
-- Efficient: Uses aggregation and filtering
SELECT
  toDate(timestamp) as date,
  count() as events
FROM events
WHERE timestamp >= today() - 30
GROUP BY date
ORDER BY date;

-- Efficient: Uses sampling for exploration
SELECT *
FROM large_table
SAMPLE 0.01  -- 1% sample
LIMIT 1000;
```

## Troubleshooting

### Connection Timeout

ClickHouse queries on large datasets can take time. Increase the timeout in your connection settings.

### Memory Limit Exceeded

Your query may be too resource-intensive. Add `LIMIT` or use sampling.