Create a Grafana dashboard(s) to present metrics generated by a SpinApp and Spin Operator

[calebschoepp]

Things like:

• requests received
• CPU consumption
• memory consumption
• egress

[calebschoepp]

CC @macolso

@rajatjindal You have any updates on this?

[rajatjindal]

Hi Caleb,

I have the dashboard to show following metrics today:

• cpu/memory usage
• replica count
• network i/o

I am not sure how to fetch # of requests metrics right now. will need to explore this more.

Also, one of the thing we are doing in the dashboard today is to use deployment metrics to fetch the name of the spin-app. It would be nice to fetch those metrics automatically from spin-operator metrics. so I'll try to test that out and submit the PR this week.

[calebschoepp]

Thanks Rajat. Things like # of requests will be probably be dependent on fermyon/spin#2293.

[macolso]

What do folks think about splitting this into two dashboards? One focused on the overall system (replicas per SpinApp, CPU, memory, network I/O, HTTP response count by type, average/median duration of SpinApp) and the second focused on an individual SpinApp

[macolso]

Also moving this issue into the "nice to have" bucket 👍