Recommended
More Related Content
What's hot
What's hot (20)
Similar to Configuration Management: What, Why, and How?
Similar to Configuration Management: What, Why, and How? (20)
Recently uploaded
Recently uploaded (20)
Configuration Management: What, Why, and How?
- 1. Configuration Management: What, Why, and How? Joseph Karns Penn State Cloud Program 11.04.17
- 2. PENN STATE IT Overview • Introduction • What is DevOps? • What is Configuration Management? • Why should you use Configuration Management? • How do you implement Configuration Management? • How does DevOps apply to Configuration Management? • Questions and Answers
- 4. PENN STATE IT Introduction • Joseph Karns • Title: Systems Administrator for Penn State • Qualifications: o RedHat Certified Systems Administrator o 10+ years of Systems Administration Experience 5 years in a DevOps Role 4 years in a Configuration Management Role • Free Time: o Network Engineer for Non-Profits o Spending Time w/Family
- 5. PENN STATE IT WHAT IS DEVOPS?
- 6. PENN STATE IT What is DevOps? • DevOps is a shift in the roles and responsibilities. • DevOps is a set of processes, ideas, and techniques.
- 7. PENN STATE IT AQuick DevOps History Lesson: • In the not-so-distant past • The typical development team • Company growth
- 8. PENN STATE IT AQuick DevOps History Lesson: • The operations team nightmares • The fundamental shift • The end result
- 9. PENN STATE IT WHAT IS CONFIGURATION MANAGEMENT?
- 10. PENN STATE IT What is Configuration Management? • Wikipedia’s Definition • My Definition • Infrastructure as Code (IAC). • The idea behind IAC • An important shift • A key insight of Configuration Management
- 11. PENN STATE IT WHY SHOULD YOU USE CONFIGURATION MANAGEMENT?
- 12. PENN STATE IT Why use Configuration Management? • Increased efficiencies, stability and control by improving visibility and tracking. • Cost reduction. • Enhanced system reliability. • The ability to define and enforce formal policies.
- 13. PENN STATE IT Why use Configuration Management? • Greater agility and faster problem resolution. • Decreased risk and greater levels of security. • More efficient change management. • Faster restoration of service.
- 14. PENN STATE IT Why use Configuration Management? • How do you improve your visibility? o Define infrastructure as code o Have multiple layers of QA o Reduced need for hardware specialists • How do you improve your tracking o Because everything is code, you can use a very common development practice of Version Control (e.g., git, mercurial, svn). Increasedefficiencies,stabilityandcontrolbyimprovingvisibilityandtracking.
- 15. PENN STATE IT Why use Configuration Management? Fasterrestorationofservice.Knowyourdesiredstate. Kitten ™DreamWorksAnimationL.L.C. • Kittens are given names like pussinboots.psu.edu • They are unique, lovingly hand raised and cared for • When they get ill, you nurse them back to health Cattle ™TeluguGateway • Cattle are given numbers like vm0042.psu.edu • They are almost identical to other cattle • When they get ill, you get another one
- 16. PENN STATE IT HOW DO YOU IMPLEMENT CONFIGURATION MANAGEMENT?
- 17. PENN STATE IT How do you implement Configuration Management? • Choose your toolset. oPopular choices include Chef, Ansible, and Puppet oChoose a tool that is a good fit for your needs oSome considerations before making a choice Infrastructure Complexity Learning Curve
- 18. PENN STATE IT How do you implement Configuration Management? • Choose your toolset. oPopular choices include Chef, Ansible, and Puppet oChoose a tool that is a good fit for your needs oSome considerations before making a choice Infrastructure Complexity Learning Curve Cost Advanced Tooling Community and Support
- 19. PENN STATE IT How do you implement Configuration Management? Ansible Puppet Chef Script Language YAML Custom DSL based on Ruby Ruby Infrastructure Controller machine applies configuration on nodes via SSH Puppet Master synchronizes configuration on Puppet Nodes Chef Workstations push configuration to Chef Server, from which the Chef Nodes will be updated Requires specialized software for nodes No Yes Yes Provides centralized point of control No. Any computer can be a controller Yes, via Puppet Master Yes, via Chef Server Script Terminology Playbook / Roles Manifests / Modules Recipes / Cookbooks Task Execution Order Sequential Non-Sequential Sequential
- 20. PENN STATE IT HOW DOES DEVOPS APPLY TO CONFIGURATION MANAGEMENT?
- 21. PENN STATE IT How does DevOps apply to Configuration Management? • Configuration Management is centered around a toolset • A toolset requires code to function • We have Infrastructure as code (IAC) • We can utilize the agile software development cycle. o Plan: your infrastructure o Code: your infrastructure o Build: your infrastructure o Test: your infrastructure o Stage: your infrastructure o Release: your infrastructure o Configure: your infrastructure o Monitor: your infrastructure
Editor's Notes
- Introduction What is DevOps? What is Configuration Management? Why should you use it? How do you implement it? How does DevOps apply to it? Questions and Answers
- Joseph Karns Title: Systems Administrator for Penn State Qualifications: RedHat Certified Systems Administrator 10+ years of Systems Administration Experience 5 years in a DevOps Role 4 years in a Configuration Management Role Free Time: Network Engineer for Non-Profits Spending Time w/Family
- DevOps is a fundamental shift in the roles and responsibilities of two IT teams that have historically been silo’d. DevOps isn’t the name of a team or a job title or even a particular technology. Instead, it’s a set of processes, ideas, and techniques.
- 1) In the not-so-distant past, if you wanted to build a software company, you also had to manage a lot of hardware. You would set up cabinets and racks, load them up with servers, hook up wiring, install cooling, build redundant power systems, and so on. It made sense to have one team, typically called Operations (”Ops”), dedicated to managing this hardware, and a separate team, typically called Developers (”Devs”), dedicated to writing the software. 2) The typical Dev team would build an application and ”toss it over the wall” to the Ops team. It was then up to Ops to figure out how to deploy and run that application. Most of this was done manually. In part, that was unavoidable, because much of the work had to do with physically hooking up hardware (e.g., racking servers, hooking up network cables). But even the work Ops did in software, such as installing the application and its dependencies, was often done by manually executing commands on a server. 3) This works well for a while, but as the company grows, you eventually run into problems. It typically plays out like this: since releases are done manually, as the number of servers increases, releases become slow, painful, and unpredictable. The Ops team occasionally makes mistakes, so you end up with snowflake servers (”kittens”), where each one has a subtly different configuration from all the others (a problem known as configuration drift). As a result, the number of bugs increases. Developers shrug and say “It works on my machine!” Outrages and downtime become more frequent.
- 1) The Ops team, tired from their pagers going off at 3 a.m. after every release, reduce the release cadence to one per week. Then to once per month. Then once every six months. Weeks before the biannual release, teams start trying to merge all their projects together, leading to a huge mess of merge conflicts. No one can stabilize the release branch. Teams start blaming each other. Silos form. The company grinds to a halt. 2) Nowadays, a profound shift is taking place. Instead of managing their own data centers, many companies are moving to the cloud, taking advantage of services such as Amazon Web Services, Azure, and Gould Cloud. Instead of investing heavily in hardware, many Ops teams are spending all their time working on software using tools such as Chef, Puppet, Terraform, and Docker. Instead of racking servers and plugging in network cables, many sysadmins are writing code. 3) As a result, both Dev and Ops spend most of their time working on software and the distinction between the two teams is blurring. It may still make sense to have a separate Dev team responsible for the application code and an Ops team responsible for the operational code, but it’s clear that Dev and Ops need to work more closely together. This is where the DevOps movement comes from.
- Ask room for answers! Show Slide - Wikipedia defines configuration management as a systems engineering process for establishing and maintaining consistency of a product's performance, functional, and physical attributes with its requirements, design, and operational information throughout its life. - That is a great definition, but I believe configuration management is the practice of building resilient, self-healing systems, and using monitoring and alerting to catch problems that can’t be resolved automatically. - If configuration management is a practice how to we management the systems? We use a concept know as Infrastructure as Code (IAC). - The idea behind IAC is that you write and execute code to define, deploy, and update your infrastructure. - This represents an important shift in mindset where you treat all aspect of operations as software – even those aspects that represent hardware (e.g., setting up physical servers). - In fact, a key insight of Configuration Management is that you can manage almost everything in code, including servers, databases, networks, log files, application configuration, documentation, automated tests, deployment processes, and so on.
- So why should you use configuration management? Well lets talk about some of the key benefits. - Increased efficiencies, stability and control by improving visibility and tracking. - Cost reduction by having detailed knowledge of all the elements of the configuration which allows for unnecessary duplication to be avoided. - Enhanced system reliability through more rapid detection and correction of improper configurations that could negatively impact performance. - The ability to define and enforce formal policies and procedures that govern asset identification, status monitoring, and auditing.
- - Greater agility and faster problem resolution, thus giving better quality of service. - Decreased risk and greater levels of security. - More efficient change management by knowing what the prior structure is in order to design changes that do not produce new incompatibilities and/or problems. - Faster restoration of service. If you know the desired state of the configuration and how they are interrelated, recovering the live configuration will be much easier and quicker. I’d like to touch on just a couple of these key benefits and then I’ll move on to How to implement it.
- The first benefit I’d like to talk more about is “Increased efficiencies, stability and control by improving visibility and tracking”. How are these items achieved with configuration management? With Improved Visibility and Tracking. - How do you improve your visibility? With your entire infrastructure being defined as code this gives you the ability to have multiple layers of quality assurance without having the need for hardware specialist. - How do you improve your tracking Because everything is code, you can use a very common development practice of Version Control (e.g., git, mercurial, svn).
- The second benefit I’d like to talk more on is “Faster restoration of service. If you know the desired state of the configuration and how they are interrelated, recovering the live configuration will be much easier and quicker.” So I’m going to stretch this concept a little to cover one of my favorite IT analogies and then wrap it back into the key benefit. That concept is Kitten vs. Cattle. Kitten: Kittens are given names like pussinboots.psu.edu They are unique, lovingly hand raised and cared for When they get ill, you nurse them back to health Cattle: Cattle are given numbers like vm0042.psu.edu They are almost identical to other cattle When they get ill, you get another one So how does this concept apply to faster restoration of service? If your infrastructure is modeled after Cattle and you have a problem arise, you just remove the piece of infrastructure and stand up a new one and apply your desired configuration state. This allows you to restore your service faster then needing to nurse that kitten back into a healthy state.
- So I’ve talked about What is Configuration Management and Why should you use Configuration Management. Lets talk about How do you implement Configuration Management?
- If there is one point that you take away for this presentation is that while configuration management is a practice, it really boils down to a toolset that is implemented into your IT eco-system There are many Configuration Management tools available in the market, each one with a different set of features and different complexity levels. Popular choices include Chef, Ansible, and Puppet. The first challenge is to choose a tool that is a good fit for your needs. There are a few things you should take into consideration before making a choice Infrastructure Complexity - Most configuration management tools require a minimum hierarchy consisting of a controller machine and a node that will be managed by it. Puppet, for example, requires an agent application to be installed on each node, and a master application to be installed on the controller machine. Ansible, on the other hand, has a decentralized structure that doesn't require installation of additional software on the nodes, but relies on SSH to execute the provisioning tasks. For smaller projects, a simplified infrastructure might seem like a better fit, however it is important to take into consideration aspects like scalability and security, which may not be enforced by the tool. Some tools can have more components and moving parts, which might increase the complexity of your infrastructure, impacting on the learning curve and possibly increasing the overall cost of implementation. Learning Curve - As mentioned earlier in this article, CM tools provide a custom syntax, sometimes using a Domain Specific Language (DSL), and a set of features that comprise their framework for automation. As with conventional programming languages, some tools will demand a higher learning curve to be mastered. The infrastructure requirements might also influence the complexity of the tool and how quickly you will be able to see a return of investment.
- Cost - Most CM tools offer free or open source versions, with paid subscriptions for advanced features and services. Some tools will have more limitations than others, so depending on your specific needs and how your infrastructure grows, you might end up having to pay for these services. You should also consider training as a potential extra cost, not only in monetary terms, but also regarding the time that will be necessary to get your team up to speed with the tool you end up choosing. Advanced Tooling - As mentioned before, most tools offer paid services that can include support, extensions, and advanced tooling. It's important to analyse your specific needs, the size of your infrastructure and whether or not there is a need for using these services. Management panels, for instance, are a common service offered by these tools, and they can greatly facilitate the process of managing and monitoring all your servers from a central point. Even if you don't need such services just yet, consider the options for a possible future necessity. Community and Support - A strong and welcoming community can be extremely resourceful for support and for documentation, since users are typically happy to share their knowledge and their extensions (modules, plugins, and provisioning scripts) with other users. This can be helpful to speed up your learning curve and avoid extra costs with paid support or training.
- This table should give you a quick overview of the main differences between three of the most popular configuration management tools available in the market today: Ansible, Puppet, and Chef. I’m not going to cover this table, but will make the slide deck available for download.
- We talked in the beginning about What is DevOps and we covered a quick history lesson, so the last point I want to cover is How does DevOps apply to Configuration Management?
- Because Configuration Management is centered around a toolset and a toolset requires code to function We have our Infrastructure represented as code (IAC) and all of this allows us to utilize the agile software development cycle, which is defined like this. - Plan: your infrastructure (Every SysAdmin knows that planning is a required first step) - Code: your infrastructure (Convert your plan into code) - Build: your infrastructure (Use the code to build your infrastructure with your toolset) - Test: your infrastructure (Test your infrastructure…you want to push it to its breaking point so you know how much it can handle) - Stage: your infrastructure (If you know its limits, let others attempt to push it to them – this is Quality Assurance) - Release: your infrastructure (Deploy your infrastructure) - Configure: your infrastructure (Handle configuration drift as you go) - Monitor: your infrastructure (Keep an every watching eye out for problems. When one arises, cycle back to step 1) This is the DevOps lifecycle and can be used in all aspects of Configuration Management. So that covers the What, Why, and How of Configuration Management. Are there any questions?
- Are there any questions?