Tech IndustryDec 29, 2018
AmazonBeffJezos!

Shifting from SOA Software Engineer to Distributed Systems Engineer

I am a generic software engineer right now and work on writing backend services, calling them from the front end and adding some monitoring and metrics to the service. Service Oriented Architecture in short. The services I contribute to already serve millions of customers. But all that is handled for me via AWS. I have used SQS, Dynamo, S3 and Lambda functions and have a good working knowledge of them. I want to shift to a job where I write code that works in a distributed environment. Something like working at Dropbox on a service that does parallel processing and uses distributed computing to serve millions of customers. Another reference would be, ability to design scalable architecture for startups that can handle growth from 50 to 5 million users. (AWS etc. kind of already does that for you. So then what's there to learn?) PS: Please let me know if the question sounds ambiguous/nonsense and I will edit it further for clarification.

Add a comment
Amazon lolwhat Dec 29, 2018

So, are you saying you don't build distributed systems at your current job at Amazon? That seems unlikely.

Amazon looking123 Dec 29, 2018

Why is that? He might be building SOA, SOA != Distributed system.

Blizzard others Dec 29, 2018

Stupid question SOA stands for what?

Airbnb jim.hodlen Dec 29, 2018

sounds like you want to work on infra

Blizzard others Dec 29, 2018

Commit to open source projects

Salesforce Pjcehm Dec 29, 2018

Know of any ?

Blizzard others Dec 29, 2018

Ur kidding ? Apache Spark/Cassandra/Kafka/Hawk/Presto/Hadoop/Beam(more of a runner)/Druid/Airflow/Ignite As well as all the ones here - https://projects.apache.org/projects.html?category And a there are a bunch of ones out that that companies have there own spin/fork of that they welcome contribution. Uber Hoodie, Splice etc

Paperless Post AlphaAF Dec 29, 2018

If your service is in a micro service architecture, running multiple instances and does asnyc processing, you already wrote a service in a distributed system. A microservice you write and deployed on a Kubernetes cluster either managed or self hosted can scale from 5 to 5 million already. I think what you want to work on is something like the core Kubernetes system which actually does all the magic. If that's the case, you can start with one of the issues on their GitHub and send a PR

Amazon Awst Dec 29, 2018

Join our team;)

This comment was deleted by the original commenter.
Amazon BeffJezos! OP Dec 30, 2018

Why is that important here?