Today the role of DevOps has become pivotal to any software delivery life cycle.
We had a big workshop today to discuss what this service means within the context of our company. We have a shared DevOps capacity used by four of our major divisions:
- Solutions;
- Product;
- Strategic Projects;
- Operations.
Not only does this discipline span various divisions, it also need to cater for our terrestrial deployments, various cloud providers (AWS, Azure, Google Cloud) and our SaaS online platform.
We also don't have separate SysOps or DataOps divisions... so DevOps is responsible for everything.
In essence, we need SUPER HUMANS... unless everything is automated... then we need someone to simply press the GREEN BOTTON!?
This means responsibilities look something like this:
- Automation of Dev, QA and Production environments in AWS and local DC’s
- Release management (DEV/QA/INT/PROD)
- Develop scripts for data extracts, fixes and deployments
- Testing of Software services (e.g. APIs) and infrastructure during deployment
- Resolve technical PROD, QA and Dev environment issues
- Building a technical knowledgebase of solutions
- Investigate and address availability, performance and capacity issues.
- Ensures the client solution build, deployment and enhancements are delivered on time and to specification both within the initial release and in all subsequent releases.
- Manage databases (deploy, optimise, backup, configure, monitor)
- System administration (Linux and Windows servers)
- Patch and upgrade management to servers, bought products and developed products
- Configuration Management
- Functional monitoring
And the minimum requirements from a skills/software stack perspective:
- AWS products and tools (full suite)
- Python and Bash scripting
- Ansible and deploy.py familiarity
- MySQL/Aurora and MongoDB
- Jenkins experience
- Docker experience
- Wildfly and Glassfish server administration and optimisation
- Understanding of auto-scaling environments and microservices and monitoring thereof
- System monitoring tools (e.g. Graphite, New Relic, Data Dog, Mint)
- Solid Linux understanding
- Continuous Integration and Continuous Deployment
Do any of you guys have the same challenge?
I would love you get some comments & suggestions from the smart Steemians!
In my opinion, you don't have to be a super human. Just a project manager having a solid software development experience and super management skills should do the trick - But that unique skill-set is unique to find since the DevOps industry is exponentially growing.
As a consultant and DevOps professional, I teach thousands of students to become DevOps Engineer :)
Downvoting a post can decrease pending rewards and make it less visible. Common reasons:
Submit
So you agree that DevOps should include all the skills I mentioned? And then having a group of DevOps resources (ie no SysOps or DBA resources) managed by a technically minded project manager?
Downvoting a post can decrease pending rewards and make it less visible. Common reasons:
Submit