Admin UI: Contextual guidance

[jseldess]

• A page that poses user questions and where/how to use Admin UI to answer. We can probably repurpose the Deploy > Monitoring and Alerting > Use the Admin UI.
• Integrate these use questions/answers into Admin UI reference docs.

[Amruta-Ranade]

Note to self: Once I have the mapping prototype, I should ask Roko and Sales Engg for feedback. (Context: Roko's YCSB presentation)

[Amruta-Ranade]

Random thought: There's multiple-to-multiple mapping of user stories to UI graphs.
Possible solution: Update three docs:

• Use the Admin UI: Selector-sections for each graph/dashboard/debug page > all possible user stories.
• Troubleshooting using the Admin UI: User stories > all possible graphs
• Individual graph/dashboard/debug pages > Intro para = context using user stories.

[Amruta-Ranade]

Note to self: I should also explain how we calculate each metric.
For context: #3744 (comment)
#3744 (comment)

It might help to cross-reference the explanation with the existing arch/concept docs (for example: link to replication layer doc, sql layer, etc).

[Amruta-Ranade]

Discussion with Andrew: Start with the custom debug chart doc and propagate the info you get to other pages.
PR for context: #3022

[Amruta-Ranade]

For Reference:

Solarwinds:
https://www.solarwinds.com/topics/multi-dimensional-database-performance-analysis
https://database.demo.solarwinds.com/iwc/main.iwc

From the interactive demo:

AWS EC2:
https://docs.aws.amazon.com/AWSEC2/latest/UserGuide/monitoring_ec2.html
https://docs.aws.amazon.com/AWSEC2/latest/UserGuide/monitoring_get_statistics.html
https://docs.aws.amazon.com/AWSEC2/latest/UserGuide/GetSingleMetricAllDimensions.html

GCP:
https://cloud.google.com/monitoring/api/v3/metrics
https://cloud.google.com/monitoring/api/metrics_gcp

[Amruta-Ranade]

For Phase 1, focus on:

• Cluster Overview
• Hardware dashboard
• Statements + Statement Details Page
• Latency
• Data distribution matrix

[Amruta-Ranade]

Note from #3794: Some of the graphs/metrics are calculated only for the gateway node. Figure out which metrics and document them.