Investigate cpu/memory spike when installing cord-platform chart

An EC2 instance of 4c-8gb just hung up when attempting to install cord-platform. 

Cannot contact ubuntu16.04-basebuild-4c-8g (i-09cfaa7db4475a615): java.lang.InterruptedException

When installing the macro helm-chart of the cord-platform for the first time may cause a cpu/memory spike in the compute node(s) which may lead to things not getting pulled/installed properly. This helm-chart consists of:

1. etcd-operator
2. kafka
3. onos
4. xos-core
5. logging
6. nem-monitoring

So you can image that when installing on a clean node/server, that is lots of containers to download. I vaguely recall seeing this issue once right before the release of SEBA-1.0 on POD1 after we deleted all the docker images, but not sure if we ever investigated. 

NOTE this may not pertain to just the cord-platform chart, but maybe the macro charts as well. I just seen it there because thats the first one to install.