英文:
Power BI Dataflow with Incremental Refresh works, but is slow to load
问题
我需要一些关于如何在 Power BI 中使用增量刷新与数据流的建议。
我有 12 个事实表（分别记录东北地区每个州的总账分录）。我需要将它们合并成一个主表。当合并后，大约有 5000 万行数据。目前我正在使用数据流来实现这一操作。我在各个表和主表上都使用增量刷新。这似乎有效，并且每晚只需大约 10 分钟来刷新。
问题是将这些数据加载到我的数据集中至少需要一个小时。只是想知道在这里的最佳做法是什么。我是否应该在数据集上也使用增量刷新？还是应该放弃数据流，只在数据集中完成所有操作？我在线上能找到的唯一信息都是几年前的了。
（另外，工作区位于高级容量上。）
英文:
I need some advice for using Power BI dataflows with incremental refresh.
I have 12 fact tables (general ledger entries for each state in the northeast.) I need to append them together into a master table. When combined, it's about 50 million rows. Right now I'm using a dataflow to achieve this. I use incremental refresh on the individual tables AND on the master table. This seems to work and only takes about 10 minutes to refresh each night.
The problem is that loading this data into my dataset takes at least an hour. Just wondering what best practices are here. Should I be using incremental refresh on the dataset as well? Or should I ditch dataflows and just do everything in the dataset? The only info I was able to find online is a few years old.
(Also, the workspace is on a premium capacity.)
答案1
得分: 1
如果您的刷新时间较长，并且可以轻松实施增量刷新，那么应该使用增量刷新。除了使您能够仅加载已更改的数据外，它还通过并行刷新分区来加速完全刷新。
英文:
>Should I be using incremental refresh on the dataset as well?
If you have a long-running refresh and you can easily implement incremental refresh, you should. In addition to enabling you to load only changed data, it also speeds up full refreshes by refreshing partitions in parallel.
通过集体智慧和协作来改善编程学习和解决问题的方式。致力于成为全球开发者共同参与的知识库，让每个人都能够通过互相帮助和分享经验来进步。
评论