Ambari installation with the rest of hadoop ecosystems

Question

I am new in Hadoop and all related topics. i have a cluster of 3 nodes (1 master and 2 Slaves) running hadoop on Ubuntu. My question is, what should i install first, Ambari then hadoop then the rest of all ecosystems? or should i install multi nodes cluster then a bunch of ecosystems then ambari? Am really confused about this, waiting your novel answers guys.

Answer 1

1. Generate SSH keys and copy between all servers. Make sure you can SSH in both directions on all machines
2. Install a supported database by Ambari, such as Postgres or MySQL
3. Install Ambari Server
4. Use a private key in the Ambari UI, which will provision Ambari Agents on all hosts
5. Use Amabri to install Hadoop stacks.

Most of this is documented in Hortonworks documentation

Installing Ambari after installing any Hadoop tools won't work out very well because Amabri needs a specific folder structure and will have conflicts when trying to overwrite existing things