A Jira Cloud integration makes the issues and projects of one Jira Cloud site queryable as tables, next to your other catalogs. You give Dippr your site name, the email of an Atlassian account and an API token for it; Dippr reads Jira live on every query, with nothing copied or stored. The tables and their columns are fixed by Dippr. You supply three values, never a URL or a query, so the integration can only talk to your own <site>.atlassian.net address.

Before you start

  • A Jira Cloud site, such as https://acme.atlassian.net. Jira Data Center and Jira Server are not supported.
  • An Atlassian API token without scopes. Create one at id.atlassian.com under Security → API tokens, signed in as the account Dippr should read Jira as, and choose the option that creates a token with no scopes. Tokens with scopes, and tokens from an Atlassian service account, are not supported: Dippr signs in to your site address with an account email and a token.
  • The email address of that account.
Everyone who is granted this catalog sees what the account behind the token can see in Jira, whatever their own Jira permissions are. Use an account whose access matches who you will grant the catalog to.

Register the integration

From Data sources, choose Add integration, select Jira Cloud, fill in the form, then choose Test & connect. Dippr has no connection test for Jira, so the test fails; tick Register anyway, without a verified connection and choose Connect. Your token is checked by your first query instead (see Test connection). Capabilities available for this kind: federation (required). Changing Jira site later asks you to enter the API token again, because the token is sent to that address.

Test connection

Dippr cannot test a Jira connection before you save it. The console still offers Test connection and Test & connect; both fail for this kind, which is why registering needs Register anyway, without a verified connection. A wrong token shows up on your first query instead, as a named error (see Troubleshooting), never as an empty table. Dippr checks the token again at most a minute after the last successful check, so a token you revoke later is reported too. For up to a minute after a successful check, a token revoked in that time can still make a query return no rows instead of the error.

In Explore

The catalog has one schema, default, with two tables. Timestamps are in UTC. issues has every issue the account can see, newest first: projects has every project the account can see: id, key, name and project_type (all varchar). There is no column that holds a whole issue. A column mask protects the column it is set on, so a column carrying the same values another way would let a reader around it.
Open issues by project
The catalog is named after the integration (acme-jira becomes acme_jira); Explore shows the name your engine uses.

Limits

  • Read-only. Dippr never changes anything in Jira.
  • Live reads are slow for large sites. Each query pages through Jira 100 issues at a time, at most 5 requests a second per engine node. A scan stops with an error after 200 pages (20,000 issues) rather than returning a partial answer. A LIMIT stops the read as soon as enough rows have arrived, so SELECT … LIMIT n with no WHERE clause works on any size of site. With a WHERE clause the read continues until n matching rows are found, so it can still reach the 200-page limit when matches are rare or far back. A site where the account can see more than 20,000 issues cannot read all of issues yet (see the next point).
  • Filters run in Dippr, not in Jira. A WHERE clause narrows the rows after they are fetched. It does not make Jira return fewer issues, so it does not make a large read cheaper or shorter. To keep reads small, use an account that can see only the projects you need.
  • One credential. Every query uses the single account you registered. There are no per-user Jira permissions.
  • Needs a current engine. An engine created before this kind existed is refused the catalog until you apply the pending platform update for that engine. The error says so (see below).
  • What you can query is governed by your own roles and grants, not by this record.

How do I fix an error?

Equivalent dip command

Register a Jira Cloud site

Data sources overview

Which sources you can connect, and how a source becomes a catalog.

Test connection and secrets

How your credentials are stored.