Background
We’re currently facing a challenge with migrating large Okta Push Groups. Previously, one of our customers had Okta push groups with 3k+ users as members. It’s worth noting that these users were added gradually over time, not all at once. Now they want to migrate these users to a new push group structure and are attempting to push 4k+ users all at once.
Issue
We’re experiencing consistent timeouts with PATCH requests during this migration process. The operation simply cannot complete when trying to migrate all users at once.
What We’ve Tried
We understand the common recommendation would be to break this down into smaller groups, but this approach is not efficient for our customer’s use case. They have specific business requirements that make maintaining the single large group structure necessary.
Questions
- What’s the best way to migrate large push groups with 3k-4k+ users without experiencing timeouts?
- Are there any specific SCIM configuration settings we should adjust to accommodate larger group sizes?
- Is there a way to increase timeout thresholds specifically for these migration operations?
- Has anyone successfully implemented an asynchronous pattern for handling large group pushes?
- Are there any best practices for optimizing our SCIM endpoint to better handle these large-scale operations?