Spike - Application performance monitoring dashboard

[raftmsohani]

Description: As a follow up on #3075 - admin page latency issue - once solution to memory limit is to investigate separating celery worker app. In order to see the resulting impact on memory consumption, we need a KPI dashboard from logs. One available option is to create a KPI using https://logs.fr.cloud.gov/ and further investigate before and after cpu/memory usage in backend app.

Open Questions:

• What percentage of backend app is being used by celery worker?
• What is the impact of separating celery worker on memory usage?

Deliverable(s): In https://logs.fr.cloud.gov/:

• [x] Create a query for cpu and memory usage by prod backend app
• [x] Can we create a dashboard with graphics for showing memory and cpu usage
• [ ] Investigate if we can export this data in CSV format?

Supporting Documentation: Please include any relevant log snippets/files/screen shots

• Cloud.gov logs

[raftmsohani]

The dashboard have been created.

Unfortunately, it is not possible to export the logs unless we use a very granular filter since CSV export has size limit.