Schego start learning about DevOps?

Hey guys!
Help with the direction I'm sysadmin for 10 years experience with UNIX, (the squid, nginx and trivia)
where to begin learning about DevOps trend?
1)I do not know how to deliver code in production (staging, etc.)
2)how to produce and deploy than do it

Anyway where to start and where to go
Please do not pingat for lubowski question and understanding
Thank you
March 12th 20 at 08:08
8 answers
March 12th 20 at 08:10
Solution
@Leonard good answer(that does not negate the fact that all the other respondents are right, too)
Devops is practice. This is not a set of tools( tools are used at certain stages, the implementation of which is necessary for the approximation to the ideal), however, certain necessary tools again still there.
About devops you can read a lot of information, but I as admin (win-admin :D) see the situation for you, as well as any, with optovoi basis, like this:
  1. Fat you have identified the questions which you here right now don't play the slightest role. @leland.Wes and @Leonard wrote why: depends on the project.
  2. Testament any ops-a: automate everything you can
    If the choice is between configuration management (chef, ansible, puppet, etc) and scripts - it is better than the first. Although there is debatable, in my project chef-th automated is not very used in the last stage of delivery in prod, since we are still printing the car up and running in AWS asg without post-configuration. Here it is possible to blue in the face arguing this is good or not, but the scripts in a perfect world lose DSL
  3. You write code to automate
    You will need git (which pulls a git hosting: bitbucket, github, networks , etc). and skills to work correctly with git. For tracking and planning changes - will need some task tracker (jira, task tracker built into networks, something else).
  4. Infrastructure as code
    Automate all means deployment automation infrastructure
    Here already come into effect the features of your environment - in the cloud you likely going to want to use terraform or, for example, CloudFormation in the AWS - built-in orchestration, or will all at once start in containers - docker , kubernetes using appropriate tools.
  5. Monitoring
    Without correct and suitable to your product monitoring(+logging) to live. And this was before DevOps trend is a classic administration. Nothing here will advise with Zabbix-Ohm himself did not get along, moved on influx and attached (TICK stack). For logging - graylog, ELK. In some parts of the used prometheus which including for Kubera comfortable. In General what will be good friends.


It's just something that you should already know and it will be easier for them to start.
Will be good to navigate (not only know, but also to choose the right tools to be able to build the right pipeline) - you would kill to have, despite the fact that you can be weak in other areas (testing, build, etc.).

For the sake of understanding cycles, you can look at (the picture is relatively random,there are many, two years ago, I focused on another, more fitting with I a list of tools, but I can not find =( )
Slide1.jpeg

P. S. I want to reiterate that the above is based on personal experience and that of moving to devops from the ops. There are those who immediately try to build all on devops parallel study optovoi part and development( seen, not at all out ). There are those who are moving in devops from the Dev. Everyone will have different opinions of what is important in order to begin
The silver Fund! Thank you for such a complete answer. - kiley commented on March 12th 20 at 08:13
March 12th 20 at 08:12
Solution
1. for a start, git plus some kind of control system (suitable familiarity with github)

2. Second CI Tool - Jenkins or Teamcity.
First, try to install a local Jenkins, and some objectives moved from cron to Jenkins, to see how it works

3. The virtual machine should not be a problem, vmware, and virtualbox is now slowly replaced by clouds and containers. That is, it would be nice to have a basic understanding of how to work with docker, aws, kubernetes. This will be enough to come assistant devops.

4. Deployment is a separate issue and depends on the project. A lot of them.

5. Would be nice to know about what makes the maven - it helps in java projects.
March 12th 20 at 08:14
Basis:
The Twelve-Factor App (transfer)

A list of required tools and technologies:
DevOps Roadmap
March 12th 20 at 08:16
Continuous Integration and Delivery - Networks TravisCI CircleCI
Docker -> docker-compose -> Kubernetes
This is a basic beginning that will deliver a sea of trouble .
March 12th 20 at 08:18
The need for knowledge bash (scripting), is a big plus knowledge of any JAPANESE, for example python. Often it will be necessary to understand someone else's code in an unfamiliar JAP.
Knowledge of Jenkins, and, for example, GitlabCI.
Any system orkestrovannaja: ansible, puppet, chief, solt...
Personally, we deploy via capistrano, but the deploy-of systcm as well previlikim set.
Each team will have its own stack of tools and techniques. You need to come to a team and jump into this maelstrom.
March 12th 20 at 08:20
March 12th 20 at 08:22
Docker, Ansible, Puppet, Chief to start. About bash and Python probably is clear.
But stack all companies are different.
March 12th 20 at 08:24

Find more questions by tags DevOpsSystem administration