hit-lacus commented on a change in pull request #1182:
URL: https://github.com/apache/kylin/pull/1182#discussion_r419248811
##########
File path: kubernetes/docker/hadoop-client/README.md
##########
@@ -0,0 +1,18 @@
+## Background
+### What is hadoop-client docker image and why we need this?
+As we all know, the node you want to deploy Kylin, should has provided Hadoop
+dependency(jars and configuration files), these dependency let you have access
+ to Hadoop Service, such as HDFS, HBase, Hive, which are needed by Apache
Kylin.
+Unfortunately, each Hadoop distribution(CHD or HDP etc.) has its own specific
jars. So, we
+can provided specific image for specific Hadoop distribution, which will make
image management task
+more easier. This will have following two benefits:
+1. Someone who has better knowledge on Hadoop can do this work, and let kylin
+ user build their Kylin image base on provided Hadoop-Client image.
+2. Upgrade Kylin will be much easier.
+
+### Build Step for CDH5.7
+1. Place Spark binary(*spark-2.3.2-bin-hadoop2.7.tgz*) into dir
`provided-binary`.
+2. Run `build-image.sh` to build image.
+
+> If you are using other hadoop distribution, please consider build image
yourself.
+
Review comment:
OK, I will add it to readme file.
----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
For queries about this service, please contact Infrastructure at:
[email protected]
