读取历史数据 [英] Reading historical data

查看:92
本文介绍了读取历史数据的处理方法,对大家解决问题具有一定的参考价值,需要的朋友们下面随着小编来一起学习吧!

问题描述

5月和6月,我们与其他Stream Analytics一起进行了概念验证.在此POC中,我们尝试从Blob中读取历史数据.由于延迟到达设置,通常您不能读取超过20天的数据.但是,在文档中 提到过,当您将延迟到达设置设为-1时,它的作用是不确定的.您不能使用Stream Analytics GUI进行设置,但是在使用ARM模板时可以设置该值.我们使用了20天以上的数据进行了测试 (准确地说是一个月又几天).并且成功导入和处理了它们(使用带有翻转窗口的分组并在此窗口内取平均值).

In may and june we did a proof of concept with among other Stream Analytics. In this POC we tried to read historical data from a Blob. Normally you can't read data older than 20 days because of the late arrival setting. However, in the documentation it is mentioned that, when you set the late arrival setting to -1 that it functions as indefinite. You can not set this by using the Stream Analytics GUI, but when using an ARM template setting this value is possible. We tested this with data older than 20 days (to be exact a month and a few days).  And they were imported and processed successfully (using a group by with a tumbling window and averages within this window).

7月,我们开始实施一个新平台,在该平台中,我们将Stream Analytics用于实时数据的获取,也用于历史数据.现在,在POC发布后的4个月内,当尝试读取历史数据时,时间超过20天,则不会发生任何输出事件 生成,也不会收到延迟到达事件.但是输入事件的数量一直在增加,没有任何输出事件.

In july we started implementing a new platform in which we use Stream Analytics for the ingestion of real time data, but also for historical data. Now, 4 months after the POC, when trying to read historical data, older than 20 days, no output events are generated, nor do we get late arrival events. But the amount of input events keeps increasing, without any output events.

旧数据的处理方式是否有所改变?

Has something changed in the way old data are processed?


推荐答案

能否让我们知道您所引用的文档的链接?
Can you please let us know the link of the document that you are referring?


这篇关于读取历史数据的文章就介绍到这了,希望我们推荐的答案对大家有所帮助,也希望大家多多支持IT屋!

查看全文
登录 关闭
扫码关注1秒登录
发送“验证码”获取 | 15天全站免登陆