1

I've noticed when retrieving a story/defect after first updating it, sometimes the retrieve response returns the field values as if the update never happened. Retrying the retrieve after a short delay (~500ms) returns the updated field values as expected. Is this a known behaviour? Is there any way of avoiding this?

I'm using the Rally API 2.0 - https://rally1.rallydev.com/slm/webservice/v2.0/

The update is being performed using this URI:

POST /slm/webservice/v2.0/Defect/14173461229?key=<key> HTTP/1.1

I'm retrieving the story after update as follows:

GET /slm/webservice/v2.0/artifact?query=(ObjectId%20=%2014173461229)&start=1&pagesize=20&fetch=true HTTP/1.1
4

1 回答 1

0

您的集成在做什么,它需要在发布更新后的 < 1 秒内重新轮询工件?是否存在第二个轮询过程来揭示更新的延迟?您的集成是否运行多个线程?响应时间是否会因一天中的时间等而有所不同?这里有许多因素可以发挥作用,但考虑到 HTTP/S 的延迟以及服务器端数据库和缓存更新等因素,500 毫秒似乎并不是一个不合理的刷新率。也就是说,为了深入了解您可能希望与Rally 支持(rallysupport@rallydev.com) 联系,因为他们有工具可以帮助评估与特定用户 ID 的请求相对应的服务器端响应时间。

于 2013-09-20T01:50:46.797 回答