zendesk
The Zendesk Support source reads users, tickets, organizations, groups, ticket fields, brands, custom roles, satisfaction ratings, tags, and ticket comments from the Zendesk Support REST API. It uses an HTTP manifest. This connector is in preview. It targets API version v2.
Configuration
Connection testing calls
GET /api/v2/users.json?page[size]=100. Requests target https://{subdomain}.zendesk.com.
Resources
All resources exceptcustom_roles, satisfaction_ratings, and ticket_comments are selected by default. custom_roles requires a Zendesk Enterprise plan or higher, and satisfaction_ratings requires administrator credentials. Selecting ticket_comments fans out across parent tickets.
Modes
All resources perform full reads. Use full upsert to refresh rows or full replace to remove rows that are no longer returned by Zendesk. When using full replace, consider these Zendesk API boundaries:- Archived tickets. Zendesk automatically archives tickets approximately 120 days after closure (three days on sandboxes, or earlier on high-volume accounts) and excludes them from
GET /api/v2/tickets.json. A full replace removes archived tickets and their comments from destination storage. - Tag limits. Zendesk lists at most the 20,000 most popular tags from the preceding 60 days. A full replace drops older or less frequent tags that fall outside this window.
Behavior
- Auth. Every request sends either HTTP Basic Authentication with
{email}/tokenandapi_token, or a Bearer token whenauth_methodisoauth_access_token. - Pagination. Endpoints use cursor-based pagination with
page[size]=100, walkingpage[after]frommeta.after_cursoruntilmeta.has_moreis false. - Rate limiting. Zendesk rate limits vary by subscription tier. Team plans allow 200 requests per minute, Growth allows 400, and Professional and Enterprise allow 700 or more. The connector paces requests at 3 requests per second with a burst allowance of 5 requests so smaller plan tiers do not experience throttling.