Background: We are building an enterprise IAM provisioning system that needs to create Jira issues at scale (100K–500K per run).
What we observed:
Single API (POST /rest/api/2/issue) — 100K calls, 1 ticket each:
- 8 concurrent workers, ~5 tickets/second
- Zero rate limits hit ✓
- Completes 100K in ~5.5 hours
Bulk API (POST /rest/api/2/issue/bulk) — 2,000 calls, 50 tickets each:
- Same 8 workers
- Hits
jira-cost-based 429 after ~1,500 tickets ✗ - Much slower due to constant 3-minute wait cycles
The paradox: Bulk API makes 50x FEWER API calls (2,000 vs 100,000) but hits rate limits much faster.
Questions:
- Why does bulk API cost significantly more per ticket than single API on the same account?
- Is
jira-cost-based limit specific to the bulk endpoint only? - Is there any way to create 100K issues in minutes (not hours) without hitting rate limits?
- Does upgrading from Standard → Premium/Enterprise plan increase the cost-based limit enough to handle 100K bulk in one run?
- Is there a provisioning-specific API or approach recommended for enterprise-scale bulk issue creation?
Account: issquaredinc.atlassian.net (Standard/Trial plan)
Headers from 429 response:
Ratelimit-Reason: jira-cost-based
X-Ratelimit-Limit: 300000
X-Ratelimit-Remaining: 0
ResetAt: [next hour]
