What are the JIRA scaling issues when we push thousands of test executions (via ZAPI) to JIRA?

[njavagal]

What are the JIRA scaling issues when we push thousands of test executions (via ZAPI) to JIRA? 

We plan to push around 20000 (20K) test executions to JIRA via ZAPI. Will the lucene index behind JIRA scale? What are some of the best practices here?

[Daniel]

Just to chime in. From testing, you would be looking at about ~7 minutes per thousand executions. During this time, you would expect the CPU load to reach very high thresholds but the memory requirements will only reach roughly 3GB during the process. These numbers include other services and JIRA itself on the machine.

[njavagal]

I would like to discuss this feature with you. What would be a good time to talk?

[MattS]

I've discussed the idea of uploading a JSON file containing multiple execution results to improve the overall duration. What do you think to the idea of adding such a feature to ZAPI?

[MattS]

Each of the 20K operations would update a custom field (build version) and set the result of a test (Pass/Fail/etc). How long would the whole 20K typically take, and what would the impact be on other JIRA users?