Title: Expansion of Storage Monitoring in the Operate First Environment
Co-op: Bryan Montalvan
School: University of Massachussetts Lowell

My project for this spring period is to expand the way we currently monitor storage metrics in the Operate First Environment. This will be done by creating a grafana dashboard which will visualize storage data by cluster and its respective namespace.

Whether you are a user or administrator, having the ability to view your cluster's storage metrics and logs is a great asset to understand the health of your application's deployment. Knowing if your PV is about to max out on resources is crucial on knowing when to request more storage. Another example is finding out what is using the most resources on your namespace.

In conjuction to this project idea I also have a complimentary project idea which is to create a Predictable Consumption Model witch can predict when a PV will be full. And could forecast other storage reated metrics/occurences. This idea can potentially involve alerting, more research will be done on this idea down the line.

First I will check the data we pull from Prometheus checking any labels/queries which are storage related and get an understanding of what type of metrics I have to my disposal. Afterwards I'll only take the metrics that users would deem useful, and the planning phase begins.

Note: When creating data visualization its always important to follow these guidelines

• Make the metrics/data make sense
• Each time a metric is pulled, it acquires a cost so pull only what you need
• Understand your intended audience

Once I have a starting idea as to what should be visualized I'll use the visulization tool, Grafana, to create visulizations which drive my projects intentions. Currently I will expand our Storage Cluster Overview dashboard which is meant to have a high level view of each cluster and its storage data. I will also create a new folder which will house a user specific dashboard so users can know more about how their projects are doing in our clusters

flowchart LR;
  OpenShift-Cluster --> Prometheus
  Prometheus --> Thanos
  Prometheus --> Grafana
  Thanos --> Grafana

(N/A)

If you are unfamilar with Operate First, please take a look at a simple introductary doc to better understand what the initative is and how the monitor infrastructure works.