How does Jira Cloud ensure data consistency between issue updates and JQL search results?

AGASTYA SHARMA

I understand that Jira Cloud uses a search index for JQL instead of querying the primary database directly. After an issue is updated through Automation, the REST API, or a bulk operation, there can sometimes be a short delay before the changes appear in JQL search results.

 

I'm curious about how this works internally.

 

1. Is the indexing process asynchronous?

 

2. Are there any documented consistency guarantees?

 

3. What is the recommended approach for apps or integrations that need to account for this delay?

 

I'd appreciate any insights or documentation explaining how Jira Cloud handles indexing and search consistency.

Gor Greyan

Hi @AGASTYA SHARMA

Welcome to the Atlassian Community!

Yes, indexing is asynchronous. As a result, Jira Cloud does not guarantee immediate read-after-write consistency for JQL searches. After an issue is created or updated, search results may temporarily return stale data until the index is updated. In most cases this takes only a few seconds, although larger operations (such as bulk edits or issue reordering) can take longer.

Atlassian's developer documentation states that the JQL Search API does not provide read-after-write consistency by default. For applications that require stronger consistency immediately after updating issues, the enhanced JQL Search API supports the reconcileIssues parameter, which ensures consistent results for the specified issue IDs (up to 50 issues).

For integrations, Atlassian recommends designing for eventual consistency. If your application updates an issue and immediately performs a JQL search, you should either use the reconcileIssues parameter when calling the enhanced JQL Search API, or implement a retry/backoff strategy to allow time for the search index to catch up.

See this documentation.
https://developer.atlassian.com/cloud/jira/platform/search-and-reconcile/