taking in node resources in account

[mudverma]

solving #513

[chaitanyaenr]

@mudverma the changes are looking good. Can we rebase the PR and also add documentation on how we compare the threshold against node capacity if limits are not set please?

[chaitanyaenr]

@mudverma tried the PR on an application with no limits set: https://github.com/cloud-bulldozer/performance-dashboards/blob/master/dittybopper/README.md#deploy-grafana-on-openshift-cluster-with-dashboards with 0 thresholds set in the config and the results includes only memory and pvc scenarios instead of all:

"recommendations": {
    "mem_services_recommendations": {
        "services": [
            "dittybopper-65fdfb476c-snfff"
        ],
        "tests": [
            "node_memory_hog",
            "pvc_disk_fill"
        ]
    }
}

}

Thoughts?

[chaitanyaenr]

LGTM

[chaitanyaenr]

@mudverma looks like there's a conflict, we might want to rebase with the latest changes. Thoughts?