Added a Grafna dashboard to give all container metrics in a single place

[GunjanJuyal]

Added a Grafna dashboard to give all container metrics (memory, CPU, I/O) in a single place. Build on top of existing cadvisor metrics.

Problem statement:

• It can be cumbersome to set up a perf testing rig, including a useful Ops dashboard for a perf-testing rig
• Given that perfiz is intended to fill this gap it will be helpful to include an out-of-box dashboard to visualize and monitor various key container metrics

Proposed solution:

• A dashboard built on top of existing cadvisor metrics, using existing Grafna data source
• Gives various gauges and charts for memory, CPU, container metrics and I/O (disk, network etc)

Possible limitations:

• In our local setup we are using a mix of docker-compose and kubernetes to set up our environment
• This leads to some variance in metrics to identify the various containers, as well as a lot of 'noise' of internal k8s containers (20+) when trying to visualize them
• The current dashboard solves this problem by splitting all charts in two lanes - one for docker-managed containers and another for k8s ones
• This may be too specific for our setup and may not be relevant for a generic demo

Possible workaround for this limitation:

• Change the dashboard to show all containers without any split
• Or, have two different dashboards - one for k8s containers and other for non-k8s ones

PFA screenshots of the existing dashboard. This will automatically show up if the current patch is applied, followed by perfiz reset / start on the demo project.

[harikrishnan83]

@GunjanJuyal Thanks for your PR, This dashboard looks great. I am testing your changes. Will update shortly on this thread.