The Project
The requests made by the client:
- - GPU Utilization
- - Cloud Costs
- - Waiting time during CI/CD
- + Queuing
- + Automation
- + Tests
- + Security
- + Monitoring
The Challenge
All the tricky technical details that bug us:
- On-demand CI/CD: Jenkins stands up running for months and costs way too much according to pace of Integration/Deployment shown by the company. A solution would be to build a role via Ansible for AWS Provision and plug it to Jenkins. That would reduce costs by a double digit number at least. Another solution implies scanning cloud costs, going through each via an automated python script, and define what could be reduced/removed.
- Jenkins Scalp: Go through Pipeline, Determine Imperative/Declarative, Implement Parallel CI, Define Delivery, AquaSec.
- Testing: Help and add UI testing.
- Handle Machine: Look through GPU use, add Queuing and complete system automation.
- Monitoring: Prometheus and Grafana to the rescue.
The Stack
All these technologies were used for this job:
