Troubleshooting Data Exchange Crashes Due to Search Service Overload

Overview

This article outlines steps to troubleshoot and resolve Jama Connect search service crashes when using the Data Exchange (DX) tool to perform imports.

1. Identifying the Issue

✔ Symptoms:

• The search service crashes when importing data via DX.

• The application becomes unresponsive or crashes entirely.

• Throttle rate adjustments impact performance:

  • High throttle rate (e.g., 10 or more requests/sec) → causes crashes.

  • Low throttle rate → leads to slow performance.

✔ Possible Causes:

• Server overload due to high request rate.

• Insufficient hardware resources (CPU, RAM).

• Elasticsearch settings misconfigured.

2. Recommended Solutions

🚨 Option 1: Reduce Request Rate in Data Exchange

✔ Set the throttle rate lower (e.g., 5 requests/sec) to prevent search service overload.
✔ Slower request rates prevent crashes, though they may increase processing time.

🚨 Option 2: Increase Server Hardware Resources

✔ Upgrade server hardware to handle higher API request rates.
✔ Recommended Memory Allocation (based on Elasticsearch settings):

• CPU: Adjust for higher compute power if experiencing slow searches.

• Memory: Increase allocated RAM based on Jama’s Elasticsearch requirements.

• Typical required memory: At least 16GB RAM, but confirm with SaaS Ops team for specifics.

🚨 Option 3: Wait for Performance Updates in Jama Connect

✔ Jama’s development team is working on performance and scalability improvements.
✔ If upgrading hardware is not feasible, monitor release notes for updates that improve search performance.

3. Best Practices for Managing Search Performance in Jama

✅ Adjust throttle rate dynamically—start low and increase gradually to find a balance.
✅ Monitor system resource utilization to determine if CPU/RAM limits are being hit.
✅ Check Elasticsearch settings and optimize search indexing configurations.
✅ Coordinate with SaaS Ops or Jama Admin Team for performance tuning guidance.

4. Conclusion

✔ The Jama search service crashes due to high API request rates from DX.
✔ Reducing the throttle rate prevents crashes but slows imports.
✔ Increasing server memory & CPU can improve performance.
✔ Future Jama updates may address search service scalability.