export-gcp-metrics: A Python repository from garg02

Export GCP MQL queries to CSV

This solution uses the Google Cloud Monitoring Query Language (MQL) to export queries to CSV.

Before starting

Ensure you have access to GCP Monitoring
Have a project with active resources and metrics (i.e. usage within the last 25 hours).

Get Code and Configure MQL Queries

Clone the repository and enter the directory:

    git clone https://github.com/garg02/export-gcp-metrics.git
    cd gcp-custom-metric-export

Add and update the MQL queries as needed in the config-template.py file. You can find examples in the MQL documentation. This repository's default queries align CPU/GPU usage with an instance's batch number over time as created in this project.

Set up environment variables

Export PROJECT_ID, START_TIME, and END_TIME variables to the shell environment. The start and end times are written in d'YYYY/MM/DD-HH:MM:SSZ' format, where Z is the time offset from UTC. Examples in Pacific time given below.

    export PROJECT_ID=`gcloud config get-value core/project`
    export START_TIME=d'2022/07/26-00:00:00-08:00'
    export END_TIME=d'2022/07/27-00:00:00-08:00'

Create configuration

Update the variables in the configuration file:
```
envsubst < config-template.py > config.py
```
If envsubst is not installed, manually open the config-template.py file, update the variables, and save it as config.py.

Export metrics to csv

Run the script
```
python3 export_metric_data.py
```