Read tables and query results from your Databricks workspace in Axiamatic.
Overview
The Databricks integration is a read-only connection that lets Axiamatic run SQL against the tables and queries you choose in your Databricks workspace. Axiamatic treats Databricks as a data lake source and labels every row it reads as issue tracking data.
Axiamatic connects through a SQL warehouse with a personal access token. The token acts as the Databricks user or service principal that owns it, so Axiamatic reads only what that identity can read.
Prepare your Databricks workspace
Complete these steps before you create the access token. The token can read only what its owner can read, so the owner needs the right permissions first.
Prerequisites: You need the Enterprise Admin or Integrations Admin role in Axiamatic. In Databricks, you need workspace admin access so you can grant entitlements and permissions.
Choose the identity that will own the token. Axiamatic recommends a service principal or a dedicated integration user because a personal token stops working when its owner changes roles or leaves the company. Prerequisite for steps 2 to 4: Each of those steps grants access to this identity.
Give the identity the Databricks SQL entitlement. Without it, the identity cannot run SQL.
Give the identity CAN USE permission on the SQL warehouse Axiamatic should use. Axiamatic runs every query on this warehouse.
Grant the identity read access to every table or view Axiamatic should read. Use Unity Catalog or table ACLs. Read-only access is enough, and Axiamatic recommends granting no write privileges.
If your workspace uses IP access lists, add the Axiamatic production address
35.167.56.23/32to the allowed list. Without it, Databricks blocks requests from Axiamatic. Skip this step if your workspace does not use IP access lists.
Create a personal access token
To let Axiamatic sign in to Databricks, generate a personal access token as the identity you prepared. In Databricks, open your user Settings, go to Developer and generate a new access token with the sql scope. Axiamatic cannot run queries with a token that lacks this scope.
When you create the token, note its expiry date in YYYY-MM-DD format. Axiamatic asks for this date during setup. Copy the token right away because Databricks shows it only once.
Store the token in a password manager or another secure place. Do not paste it into tickets, chat threads or shared documents.
Find your connection details
Axiamatic needs your workspace URL and the HTTP path of your SQL warehouse. The last part of the HTTP path is the warehouse ID, which Axiamatic uses to run every query.
To find your workspace URL, sign in to Databricks in a browser and copy the address up to the end of the domain. It looks like https://dbc-1234abcd-5678.cloud.databricks.com on AWS or https://adb-1234567890123456.7.azuredatabricks.net on Azure. Leave out any trailing slash or path.
To find the HTTP path, open SQL Warehouses under SQL in the sidebar and select the warehouse Axiamatic should use. Open the Connection details tab and copy the HTTP path. For example, in /sql/1.0/warehouses/a1b2c3d4e5f6a7b8 the warehouse ID is a1b2c3d4e5f6a7b8.
Connect Databricks to Axiamatic
Complete the sections above before you start. You need your workspace URL, the access token and the token's expiry date. You also need the Enterprise Admin or Integrations Admin role in Axiamatic.
Click your profile menu and select Program Settings.
Open the Integrations page. In the Data & Analytics category, find the Databricks card.
Click Connect. A window opens and asks what type of connection you want.
Choose how to scope the connection, then click Proceed.
Shared makes the connection available to all current and future projects and business units in your workspace.
Private limits the connection to the current project. Other projects need their own connection.
Axiamatic recommends Shared because you set it up once and every project can use the same Databricks data. Choose Private if your organization restricts which projects can see this data.In the Linking your Databricks account window, enter your details.
Expiry Date: the token's expiry date in
YYYY-MM-DDformat. This field is required.Host: your workspace URL, such as
https://dbc-1234abcd-5678.cloud.databricks.com.Access Token: the personal access token you created.
Click Connect.
After you click Connect, Axiamatic checks your details and the Databricks card shows the connection status. The integration is then ready for you to choose the tables or queries Axiamatic syncs.
Choose tables or queries
You can share a table or a SQL query with Axiamatic. Axiamatic recommends one table or query per connection.
To share a table, give its fully qualified name in the form catalog.schema.table, such as workspace.support.tickets. Axiamatic uses the table name as the export label unless you set another one. To read only rows that changed since the last run, choose a date or timestamp column such as updated_at. Without one, every run exports the whole table.
To share a query, give it a name and the SQL text. Axiamatic uses the query name as the export label. To read only changed rows, put UPDATED_TIME_STAMP_AXM_MILLIS_PLACEHOLDER (epoch milliseconds) or UPDATED_TIME_STAMP_AXM_PLACEHOLDER (a timestamp) where the last-run time belongs. Axiamatic replaces the placeholder with the time of the last run. Without a placeholder, every run exports everything.
Add an ORDER BY on the update column and do not end the SQL with a semicolon. For example: SELECT * FROM workspace.support.tickets WHERE updated_at > UPDATED_TIME_STAMP_AXM_MILLIS_PLACEHOLDER ORDER BY updated_at
Frequently Asked Questions
Does Axiamatic write to Databricks?
Does Axiamatic write to Databricks?
No. Axiamatic only runs read-only SQL, so the token owner needs read access and no write privileges.
Does this work with Azure Databricks?
Does this work with Azure Databricks?
Yes. Axiamatic connects to Databricks workspaces on AWS and Azure.
Why can't Axiamatic see one of my tables?
Why can't Axiamatic see one of my tables?
The token owner probably lacks read access to it. Grant read access to that table or view in Unity Catalog or through table ACLs.
Why does every run export the whole table?
Why does every run export the whole table?
The table has no incremental column, or the query has no timestamp placeholder. Add one to read only the rows that changed since the last run.
Why does Databricks reject requests from Axiamatic?
Why does Databricks reject requests from Axiamatic?
Your workspace likely uses an IP access list that does not include Axiamatic. Add 35.167.56.23/32 to the allowed list.
For pricing information, demo requests, or assistance, contact us at [email protected]




