@revmischa - Amazon Web Services Thought Lord
Who Am I?
I am a software engineering professional based in San Francisco. I have a wealth of experience with software development and operations, and one of my specialties is AWS architecture and provisioning. I’m advertising my services as a consultant on designing, implementing, automating, improving and saving time and money with AWS.
What Have I Done?
- Live-migrated a health-tech company’s infrastructure from a traditional datacenter to a HIPAA-compliant, secure AWS environment.
- Implemented functionality with EC2, S3, SNS, SQS, Lambda, CloudWatch, CloudFront, ElasticCache, API Gateway, RDS PostgreSQL, Route53, CodeDeploy, IAM, IoT, SES.
- Released a command-line tool suite named “Unemployed DevOps” for wrangling clusters with AutoScaling Groups and CodeDeploy.
- Presented a talk on AWS about best practices, automating everything and how to simplify your setup using standard unix tools that have been around for decades.
- Designed monitoring and alerting systems based around CloudWatch, Lambda, SQS, Slack.
- Reduced costs by eliminating wasteful services, using AutoScaling, removing special snowflake instances, optimizing request flow and server load to take full advantage of EC2 instances.
- Massively streamlined deployments with AWS-native CodeDeploy.
- Reduced web application server load and optimized static file delivery with nginx and CloudFront CDN.
- Created highly available web application services and RDS multi-AZ database failover.
Do I Know What I’m Doing?
I do not claim to know everything about AWS, but I do have access to many people who are highly knowledgeable. When I need advice or have questions I often speak with Amazon directly either via support or in-person at their AWS Loft in San Francisco. They have guided me and validated my general approach and specific deployments.
What Can I Do For You?
Review your infrastructure, cut down costs, validate security models, make recommendations on efficiency improvements, and most importantly, automate the heck out of everything. I am not seeking long-term employment doing AWS management. You should have everything automated anyway. I will help you get there. I am first and foremost a lazy engineer.
How Do I Do It?
- Automate common tasks like deployment, reading tags, SSHing into hosts and much more using awscli bash scripts or boto3 python scripts.
- Consolidate useful scripts and commands using a simple Makefile.
- Create RPMs for instance configuration, dependencies, provisioning.
- Use Udo, a simple command-line interface for defining clusters and roles for instances using AutoScaling groups. Define packages to install and generate reusable cloud-init scripts for each cluster and role.
- Create CloudWatch lambdas and alerts for monitoring and notifications.
- Consolidate disparate services into PostgreSQL, which can act as your pubsub message passing server, full-text search (still using ElasticSearch? how quaint!), JSONB for document stores and much much more. You are using PostgreSQL, right?
What I Won’t Do For you
- Configure your Oracle license
- Defend your hare-brained microservices architecture idea
- Read HackerNews comments
- Create enterprise java build systems
- Waste time with configuration management systems for AWS (seriously stop doing this it’s wrong)
- Troubleshoot your Elastic Beanstalk setup. You probably don’t need EB anyway. Use Udo and AutoScaling groups instead.
Sounds Interesting, Now What?
Hit me up and we can chat and see if I can help you out. If I can’t, no problemo.