# How do you connect to Nova Data Share?

Source: https://novadata.io/docs/data-share/connect
Checked against Nova's code on 2026-10-29

You connect to `share.novadata.io`, database `nova_share`, with the username and password Nova sends you, over TLS. Any tool with a MySQL connector works on port 3306; ClickHouse clients can use the native protocol or HTTPS with the same login.

## Which interface should you use?

All three reach the same tables with the same row filtering. TLS is required on each.

| Interface | Port | Use it when |
|---|---|---|
| MySQL wire protocol | `3306` | Your tool has a MySQL connector: Looker Studio, Power BI, Tableau, Metabase, DBeaver, Python |
| Native ClickHouse (TCP, TLS) | `9440` | `clickhouse-client`, or copying into your own ClickHouse with `remoteSecure` |
| HTTPS | `8443` | `clickhouse-connect`, `clickhouse-js`, the JDBC driver, plain HTTP |

## How do you receive your login, and how is it revoked?

Nova sends your username and password apart from the rest of the setup, through an expiring share link. The username looks like `mysql4qjne6q9i1y_yourname`: keep it whole, prefix included, on every port.

To revoke access, ask your Nova contact. Nova deletes the login itself, which closes all three ports at once.

## How do you connect a BI tool?

### Looker Studio

1. Create, Data source, then pick the **MySQL** connector by Google.
2. Choose database authentication: host `share.novadata.io`, port `3306`, database `nova_share`, your username and password.
3. Tick **Enable SSL** and upload Let's Encrypt's root certificate, ISRG Root X1 (`isrgrootx1.pem` from letsencrypt.org), as the server certificate. Leave the client certificate and key empty.
4. Authenticate, then pick a table or paste a **Custom Query**. A query that groups by day and marketplace keeps reports fast.

### Power BI, Tableau, Metabase, DBeaver

Add a MySQL source with the same host, port `3306`, database `nova_share`, your login, and SSL on.

### Command line

`mysql -h share.novadata.io -P 3306 -u mysql4qjne6q9i1y_yourname -p --ssl-mode=REQUIRED`

If a connection hangs for about two minutes and times out, a firewall on your network is usually dropping it. Allow `share.novadata.io`, which resolves to `51.159.198.29`.

## What limits apply?

| Limit | Value |
|---|---|
| Query time | 120 seconds |
| Rows returned | 50 million per query |
| Memory | 4 GB per query |
| Queries | 10,000 per hour per login |

The share is for pulling, not for live dashboards: copy the tables into your own warehouse and point reports at your copy. A daily or two-hourly pull of the last 60 days covers Amazon's restatements on sales and fees; 20 days is enough for the advertising tables. Replace current-state tables, such as `listing_health` or `advertising_negative_targets`, whole on each pull. To seed a warehouse, extract the large tables month by month.

## Which time zone are the dates in?

| Column | Meaning |
|---|---|
| `metric_date` on `product_metrics`, `purchase_day`, `posted_day` | Day in the marketplace's time zone |
| `metric_date` on the advertising tables | Day in the Ads profile's time zone |
| `snapshot_date` on `inventory` | Day Amazon generated the stock report |
| Columns ending in `_utc`, `updated_at`, `updated_through` | UTC |

The server clock is UTC, so `today()` returns the UTC date. To read one marketplace day, filter on the date value itself, for example `metric_date = '2026-10-27'`. See [Time zones](/docs/your-data/time-zones).

## What is a good first query?

Units, sales and net profit by marketplace over the last 30 days:

`SELECT marketplace_name, currency, sum(units) AS units, sum(product_sales) AS sales, sum(net_profit) AS net_profit FROM nova_share.product_metrics WHERE metric_date >= today() - 30 GROUP BY marketplace_name, currency ORDER BY sales DESC`

Every money column is already in your organization's currency and signed so that a plain `sum()` gives the total. Then read [Tables](/docs/data-share/tables) before you join or add tables together.

## Ask your AI

> Write a SQL query for my Nova Data Share that returns, per week and marketplace, units, sales, ad spend and net profit for the last 12 weeks.