Luca's suggestion is good. Starting from getting things running is a good way to start, hence provisioner.
To break down the whole landscape into stages: Step 1. Build Env Setup(this is available as docker images) Step 2. Build RPM/DEB packages (this is available by Bigtop releases repos: https://downloads.apache.org/bigtop/bigtop-1.4.0/repos/) Step 3. Provision a Hadoop cluster (you can leverage provisioner to get things running quickly) Once you have Step 3 running, you can work backwards for customization to build you own big data stack by Bigtop. Evans Joba1 <joachim.banz...@googlemail.com> 於 2020年11月1日 週日 上午1:15寫道： > Thank you Luca, very helpful. Will keep me busy for a while. > -- > iphone so i typo >
