1

I am a beginner in Apache Beam and very curious to understand the internals of Apache Beam. I read some pages and watched some videos and all are explaining how watermarks help to handle the readiness and obsolescence of an infinite stream. Basically handling late data. But no one explained how Apache Beam estimates the watermark.

Can you help me understand the basics of watermarks? How does Apache Beam estimate the watermarks? You can also point me to some docs that can help me understand the basics of this.

arora
  • 11
  • 3

1 Answers1

0

The Beam programming guide is very complete on this topic :

Beam programming guide

There are the following parts : 8.4. Watermarks and late data and 8.4.1. Managing late data

Mazlum Tosun
  • 5,761
  • 1
  • 9
  • 23