add note about sparkmaster to each section

This commit is contained in:
Khalid Huseynov 2016-11-15 23:25:20 +09:00
parent 887b499143
commit 78e7907603

View file

@ -32,9 +32,8 @@ So [install docker](https://docs.docker.com/engine/installation/) on the machine
[Spark standalone](http://spark.apache.org/docs/latest/spark-standalone.html) is a simple cluster manager included with Spark that makes it easy to set up a cluster.
You can simply set up Spark standalone environment with below steps.
> **Note 1:** Since Apache Zeppelin and Spark use same `8080` port for their web UI, you might need to change `zeppelin.server.port` in `conf/zeppelin-site.xml`.
> **Note :** Since Apache Zeppelin and Spark use same `8080` port for their web UI, you might need to change `zeppelin.server.port` in `conf/zeppelin-site.xml`.
> **Note 2:** `sparkmaster` hostname used to run docker containers below should be defined in your `/etc/hosts`.
### 1. Build Docker file
You can find docker script files under `scripts/docker/spark-cluster-managers`.
@ -57,6 +56,8 @@ docker run -it \
spark_standalone bash;
```
Note that `sparkmaster` hostname used here to run docker container should be defined in your `/etc/hosts`.
### 3. Configure Spark interpreter in Zeppelin
Set Spark master as `spark://<hostname>:7077` in Zeppelin **Interpreters** setting page.
@ -113,6 +114,8 @@ docker run -it \
spark_yarn bash;
```
Note that `sparkmaster` hostname used here to run docker container should be defined in your `/etc/hosts`.
### 3. Verify running Spark on YARN.
You can simply verify the processes of Spark and YARN are running well in Docker with below command.
@ -174,6 +177,8 @@ docker run --net=host -it \
spark_mesos bash;
```
Note that `sparkmaster` hostname used here to run docker container should be defined in your `/etc/hosts`.
### 3. Verify running Spark on Mesos.
You can simply verify the processes of Spark and Mesos are running well in Docker with below command.