Can you find “DevOps Success” by reading only IT literature?
The answer is “mostly No, but a little bit Yes”, for a number of reasons.
The main reason is that many of the blogs, whitepapers and webinars around DevOps are ultimately about technology and toolchains. Whilst they might reference the DevOps C.A.L.M.S model in passing the conversation is generally focussed on the A for Automation and the M for Metrics.
But what both of those books have in common is that they drew extensively from non-IT business literature.
Goldratt’s “The Goal” & TOC, Systems Thinking, Lean manufacturing, Deming and Kanban being major influences on the Phoenix project, and Mandi’s e-book drawing on business-centric cultural/organisational change literature.
“Searching on the Harvard Business Review website for “cultural change” will get you 60+ publications going back nearly 30 years” – Mandi Walls
“Lean Manufacturing”, “Strategic Alignment”, “Organisational Change”, “Culture”, “Business Transformation” and many other topics have been staples of the MBA curriculum in business schools for many years and there is a wealth of resources available online to explore. Reading beyond the IT literature and exploring the wider business context for your DevOps Transformation will, we believe, significantly increase your chances of getting business buy-in and having a successful outcome to your DevOps change programme.
In order to make this easier for you we (@DevOpsGuys) will be publishing a weekly blog post exploring an area of business literature and how it can be used in DevOps.
We’re call it the #DevOpsMBA :-)
So please subscribe to our blog (link on right), follow us on Twitter or search for the #DevOpsMBA hashtag on Twitter to keep informed!
Despite what some people seem to think there is more to DevOps than just Continuous Delivery and Infrastructure Automation with Puppet, Chef or Ansible.
To me, DevOps is “an alternative model for the creation of business value from the software development life-cycle that encompasses a product-centric view across the entire product life-cycle (from inception to retirement) and recognises the value in close collaboration, experimentation and rapid feedback”.
Moving from one model of value creation can either be an organic process or a transformational one – you can “grow into” the new model or you can plan a strategy of change to transform your organisation from one to the other.
It’s in this “Organic DevOps” versus “Transformational DevOps” that I see a growing disconnect between different sectors of the DevOps community, particularly between “DevOps for Start-ups” and “DevOps for Enterprise”.
IMHO, “Start-Ups DevOps” normally follows the “organic DevOps” path – you’re often starting from a relatively “greenfields” approach based on a cloud infrastructure. You probably already have a very close, collaborative culture because there’s only 20 of you anyway and you all work in the same office and you spend 18hrs a day there. Automation is part of your DNA because you’ve never had the staffing levels to do it manually.
“Enterprise DevOps” is normally “Transformational DevOps” – you have large, distributed IT teams that cross geographic locations, time-zones and probably organisational boundaries (due to outsourcing). You have extensive legacy applications and infrastructure estates (JVM 1.4 on Tomcat 5 anyone?) and you’re likely to have well developed Prince2/ITIL/SixSigma delivery models rigidly enforced by a centralised command&control mindset, backed by an army of highly-paid consultant from the Big 5 telling your CEO, CIO and CTO the best way to manage their IT budget.
Moving an enterprise to DevOps via a transformation programme is a very different challenge to introducing DevOps concepts into a receptive start-up and watching them grow organically, and the DevOps community needs to make sure that when it’s evangelising DevOps to the world that it’s aware of the differences and challenges inherent in each approach.
If you want to debate this idea of “Start-up Organic versus Enterprise Transformational DevOps” we’re taking part in a Webinar tonight with the great folks over at ScriptRock that’s focussing on Enterprise DevOps. It’s at 1900 BST, 11:00am PT / 2:00pm ET (60 minutes).
We’d really like to get your thoughts on this by asking a question on the webinar or by leaving a comment below as these concepts are still experimental and, just like DevOps itself, the faster we get feedback and the more we iterate around the concept the stronger it will be!
A thought-provoking infographic – with some interesting data points – shows how companies are reaping real rewards from investing in agile software delivery processes. Check out the graphic – from Zend – for more on how DevOps and Continuous Delivery are bridging the speed and innovation gap between business demand and IT.
Continuous Delivery Infographic by Zend Technologies.
A fascinating TED talk from “Predictably Irrational” author Dan Ariely has some interesting pointers to some of the underlying psychological mechanisms that make the DevOps model a better way to structure work within IT departments.
“we care much more about a product if we’ve participated from start to finish rather than producing a single part over and over.”
- Seeing the fruits of our labour may make us more productive
- The less appreciated we feel our work is, the more money we want to do it
- The harder a project is, the prouder we feel of it
- Knowing that our work helps others may increase our unconscious motivation
- The promise of helping others makes us more likely to follow rules
- Positive reinforcement about our abilities may increase performance
- Images that trigger positive emotions may actually help us focus
Many of these tie directly back to key DevOps principles.
For example the “First Way of DevOps” encourages “systems thinking” which relates directly to #1 above – if we are looking at the entire system (not just our small part) we will inherently be looking at the “fruits of our labour”.
Similarly fostering a team-based DevOps culture where we can see “how our work impacts on others” is closely aligned with #4.
For me, #2 and #6 tie directly back to “leadership” (as opposed to “management”). Good leaders know that praise (either private 1:1 praise with individuals or public praise in front of the team) can have a huge impact on morale, with a subsequent impact on productivity and quality.
It’s fascinating to see how behavioural science is increasing our understanding of human motivation. The challenge for us in the DevOps movement is to take these science-based insights and see how we can apply them with our teams to create a better way of working.
We put together this “DevOps Intro” to the recruitment team at a London recruitment consultancy to help them understand the DevOps Market place.
The goal was to help them understand both the WHY and the WHAT of DevOps and what that might mean for recruiters.
Hopefully you will find this interesting as well!
Join us for “The Top 10 DevOps Operational Requirements” – http://www.brighttalk.com/webcast/534/98059 via @BrightTALK
One of the key tenets in DevOps is to involve the Operations teams in the full software development life cycle (SDLC) and in particular to ensure that “operational requirements” (“OR’s”, formerly known as “non-functional requirements”, “NFR’s”) are incorporated into the design&build phases.
In order to make your life easier the DevOpsGuys have scoured the internet to compile this list of the Top Ten DevOps Operational Requirements (ok, it was really just chatting with some of the guys down the pub BUT we’ve been doing this a long time and we’re pretty sure that if you deliver on these your Ops people will be very happy indeed!).
#10 – Instrumentation
Would you drive a car with a blacked out windscreen and no speedo? No, didn’t think so, but often Operations are expected to run applications in Production in pretty much the same fashion.
Instrumenting your application with metrics and performance counters gives the Operations people a way to know what’s happening before the application drives off a cliff.
Some basic counters include things like “transactions per second” (useful for capacity) and “transaction time” (useful for performance).
#9 – Keep track of the Dependencies!
“Oh yeah, I forgot to mention that it needs [dependency XYZ] installed first” or “Yes, the system relies on [some 3rd party web service] can you just open up firewall port 666 right away”.
Look, we all understand that modern web apps rely on lots of 3rd party controls and web services – why re-invent the wheel if someone’s already done it, right? But please keep track of the dependencies and make sure that they are clearly documented (and ideally checked into source control along with your code where possible). Nothing derails live deployments like some dependency that wasn’t documented and that has to be installed/configured/whatever at the last moment. It’s a recipe for disaster.
#8 – Code defensively & degrade gracefully
Related to #9 above – don’t always assume the dependencies are present, particularly when dealing with network resources like databases or web services and even more so in Cloud environments where entire servers are known to vanish in the blink of an Amazon’s eye.
Make sure the system copes with missing dependences, logs the error and degrades gracefully should the situation arise!
#7 – Backward/Forward Compatibility
Existing code base with new database schema or stored procedure?
New code base with existing database schema or stored procedures?
Either way, forwards or backwards, it should work just fine because if it doesn’t you introduce “chicken and the egg” dependencies. What this mean for Operations is that we have to take one part of the system offline in order to upgrade the other part… and that can mean an impact on our customers and probably reams of paperwork to get it all approved.
#6 – Configurability
I once worked on a system where the database connection string was stored in a compiled resource DLL.
Every time we wanted to make a change to that connection string we had to get a developer to compile that DLL and then we had to deploy it… as opposed to simply just editing a text configuration file and re-starting the service. It was, quite frankly, a PITA.
Where possible avoid hard-coding values into the code; they should be in external configuration files that you load (and cache) at system initialisation. This is particularly important as we move the application between environments (Dev, Test, Staging etc) and need to configure the application for each environment.
That said, I’ve seen systems that had literally thousands of configuration options and settings, most of which weren’t documented and certainly were rarely, if ever, changed. An “overly configurable” system can also create a support nightmare as tracking down which one of those settings has been misconfigured can be extremely painful!
#5 – “Feature Flags”
A special case of configurability that deserves its own rule – “feature flags”.
We freakin’ love feature flags.
Because they give us a lot of control over how the application works that we can use to (1) easily back out something that isn’t working without having to roll-back the entire code base and (2) we can use it to help control performance and scalability.
#4 – Horizontal Scalability (for all tiers).
We all want the Product to be a success with customers BUT we don’t want to waste money by over-provisioning the infrastructure upfront (we also want to be able to scale up/down if we have a spiky traffic profile).
For that we need the application to support “horizontal scalability” and for that we need you to think about this when designing the application.
3 quick “For Examples”:
- Don’t tie user/session state to a particular web/application server (use a shared session state mechanism).
- Support for read-only replicas of the database (e.g. a separate connection string for “read” versus “write”)
- Support for multi-master or peer-to-peer replication (to avoid a bottleneck on a single “master” server if the application is likely to scale beyond a reasonable server specification). Think very carefully about how the data could be partitioned across servers, use of IDENTITY/@Auto_Increment columns etc.
#3 –Automation and “scriptability”
One of the key tenets in the CALMS DevOps Model is A for Automation (Culture-Automation-Lean-Metrics-Sharing if you want to know the others).
We want to automate the release process as much as possible, for example by packaging the application into versionable released or the “infrastructure-as-code” approach using tools like Puppet & Chef for the underlying “hardware”.
But this means that things need to be scriptable!
I can remember being reduced to using keystroke macros to automate the (GUI) installer of a 3rd party dependency that didn’t have any support for silent/unattended installation. It was a painful experience and a fragile solution.
When designing the solution (and choosing your dependencies) constantly ask yourself the question “Can these easily be automated for installation and configuration”? Bonus points if you can, in very large scale environments (1,000 of servers) build in “auto-discovery” mechanisms where servers automatically get assigned roles, service auto-discovery (e.g. http://curator.apache.org/curator-x-discovery/index.htm) etc.
#2 – Robust Regression Test suite
Another think we love, almost as much as “feature flags” is a decent set of regression test scripts that we can run “on-demand” to help check/verify/validate everything is running correctly in Production.
We understand that maintaining automated test scripts can be onerous and painful BUT automated testing is vital to an automation strategy – we need to be able to verify that an application has been deployed correctly, either as part of a software release or “scaling out” onto new servers, in a way that doesn’t involve laborious manual testing. Manual testing doesn’t scale!
The ideal test suite will exercise all the key parts of the application and provide helpful diagnostic messaging if something isn’t working correctly. We can combine this with the instrumentation (remember #10 above), synthetic monitoring, Application Performance Management (APM) tools (e.g. AppDynamics), infrastructure monitoring (e.g. SolarWinds) etc to create a comprehensive alerting and monitoring suite for the whole system. The goal is to ensure that we know something is wrong before the customer!
#1 – Documentation
Contrary to popular belief we (Operations people) are quite happy to RTFM.
All we ask is that you WTFM (that’s W as in WRITE!) J
Ideally we’d collaborate on the product-centric documentation using a Wiki platform like Atlassian Confluence as we think that this gives everyone the easiest and best way to create – and maintain – documentation that’s relevant to everyone.
As a minimum we want to see:
- A high-level overview of the system (the “big picture”) probably in a diagram
- Details on every dependency
- Details on every error message
- Details on every configuration option/switch/flag/key etc
- Instrumentation hooks, expected values
- Assumptions, default values, etc
Hopefully this “Top Ten” list will give you a place to start when thinking about your DevOps “Operational Requirements” but it’s by no means comprehensive or exhaustive. We’d love to get your thoughts on what you think are the key OR’s for your applications!
Our experienced DevOps team provides a fully-managed “application-centric” website support service to your business and your customers. Contact us to today, to find out how we can help.
image source: - CC - c_knaus via Flickr - http://www.flickr.com/photos/soda37/6496536471
In a previous post we talked a lot about the “Product-centric” approach to DevOps but what does this mean for the role of the Agile “Product Owner”?
So what is the traditional role of the Product Owner? Agile author Mike Cohn from MountainGoat Software defines it thus:
“The Scrum product owner is typically a project’s key stakeholder. Part of the product owner responsibilities is to have a vision of what he or she wishes to build, and convey that vision to the scrum team. This is key to successfully starting any agile software development project. The agile product owner does this in part through the product backlog, which is a prioritized features list for the product.
The product owner is commonly a lead user of the system or someone from marketing, product management or anyone with a solid understanding of users, the market place, the competition and of future trends for the domain or type of system being developed” – Mike Cohn
The definition above is very “project-centric” – the Product Owner’s role appears to be tied to the existence and duration of the project and their focus is on the delivery of “features”.
DevOps, conversely, asks us (in the “First Way of DevOps”) to use “Systems Thinking” and focus on the bigger picture (not just “feature-itis”) and the “Product-centric” approach says we need to focus on the entire lifecycle of the product, not just the delivery of a project/feature/phase.
Whilst decomposing the “big picture” into “features” is something we completely agree with, as features should be the “unit of work” for your Scrum teams or “Agile Software Development Factory”, it needs to be within the context of the Product Lifecycle (and the “feature roadmap”).
So the key shift here then is to start talking about the “Product Lifecycle Owner”, not just the Product Owner, and ensure that Systems Thinking is a critical skill for that role.
The second big shift with DevOps is that “Non-Functional Requirements” proposed by Operations as being critical to the manageability and stability of the product across its full lifecycle “from inception to retirement” must be seen as equally important as the functional requirements proposed by the traditional Product Owner role.
In fact, we’d like to ban the term “Non-Functional Requirements” (NFR’s) completely, as the name itself seems to carry an inherent “negativity” that we feel contributes to the lack of importance placed on NFR’s in many organisations.
We propose the term “Operational Requirements” (OR’s) as we feel that this conveys the correct “product lifecycle-centric” message about why these requirements are in the specification – “This is what we need to run and operate this product in Production across the product’s lifecycle in order to maximise the product’s likelihood of meeting the business objects set for it”.
We propose the term “Operational Requirements” (OR’s) as we feel that this conveys the correct “product lifecycle-centric” message about why these requirements are in the specification.
For the slightly more pessimistic or combative amongst you the “OR” in Operational Requirements can stand for “OR this doesn’t get deployed into Production…” .
The unresolved question is do we need an “Operational Product Owner” or does the role of the traditional, business-focussed Product Owner extend to encompass the operational requirements?
You could argue that the “Operational Product Owner” already partly exists as the “Service Delivery Manager” (SDM) within the ITIL framework but SDM’s rarely get involved in the software development lifecycle as they are focussed on the “delivery” part at the end of the SDLC. Their role could be extended to include driving Operational Requirements into the SDLC as part of the continual service improvement (CSI) process however.
That said, having two Product Owners might be problematic and confusing from the Agile development team perspective so it would probably be preferable if the traditional Business product owner was also responsible for the operational requirements as well as the functional requirements. This may require the Product Owner to have a significantly deeper understanding of technology and operations than previously otherwise trying to understand why “loosely-coupled session state management” is important to “horizontal scalability” might result in some blank faces!
So in summary a “DevOps Product Owner” needs to:
- Embrace “System Thinking” and focus on the “Product Lifecycle” not just projects or features
- Understand the “Operational Requirements” (and just say “No to NFR’s”!)
- Ensure that the “OR’s” are seen as important as the “Functional Requirements” in the Product roadmap and champion their implementation
In future posts we’ll examine the impact of DevOps on other key roles in the SDLC & Operations. We’ve love to get your opinions in the comments section below!
image source: – CC - CannedTuna via Flickr - http://www.flickr.com/photos/cannedtuna/7348317140/sizes/m/
When doing the research for our BrightTalk Webinar on DevOps I came across this quote from Jez Humble on “products not projects”, which really struck a chord with our thinking about what we call the “product-centric” approach to DevOps.
Figure 1 – DevOpsGuys BrightTalk Webinar
One of the key elements of DevOps is to ensure that IT strategy is directly linked with Business strategy.
One of the critical scenes in “The Phoenix Project” is where our hero Bill gets to meet with the CFO and they work through the list of “pet IT projects” in the organisation and find that many of them can’t really be tied back to any business strategy or organisational goal.
This is, we feel, one of the major problems with the “project-centric” view of IT and why we need to push for a “product-centric” view.
The “project-centric” view, whilst important for mobilising resources and organising activities, can easily get disconnected from the original business objectives. Any “Project”, just like the “Phoenix Project” in the novel, runs the risk of becoming its own “raison d’etre” as it becomes more about getting “the Project” over the line than whatever business benefits were originally proposed.
In contrast a “Product-centric” viewpoint is focussed on the Product (or Service) that you are taking to market. By keeping the focus on “the Product” you are constantly reminded that you are building a product, for customers, as part of an organisational objective that ties back to over-arching business strategy.
For example if you were in the travel sector you might be adding a new “Online Itinerary Manager” product to your website to enable your customers to view (and possibly update) their itinerary online as part of your business strategy to both empower customers online and reduce the number of “avoidable contacts” to your call centre (and hence reduce costs).
One of the other benefits of the “product-centric” view is also highlighted in Jez’s quote above – “From Inception to Retirement”.
The “product-centric” view reminds us to think about the “Product lifecycle” and not just the “software development lifecycle”.
The “product-centric” view reminds us to think about the “Product lifecycle”
and not just the “software development lifecycle”.
You need to understand how this product is going to be deployed, managed, patched, upgraded, enhanced and ultimately retired… and that means you need close cooperation between the business, the developers and operations (= DevOps!).
So how do you introduce the “Product-centric” view into an organisation that might already have existing website that offers products/services to customers?
Well, firstly, you need to stop referring to it as “the website”.
A “website” is a platform and a channel to market, it’s not a “product”.
A product is a good or service that you offer to the market in order to meet a perceived market need in the hope that you will in return received an economic reward.
For some “websites” there might be a single product e.g. Dominos sells pizza online (food products) and for other there might be multiple products e.g. theAA.com sells membership (breakdown services), Insurance, Financial Services, Driver Services (mostly training) and Travel Services. If you’re ever in doubt on the products your website sells your top navigation menu will probably give you a pretty good indication!
Figure 2 – what products does the AA sell?
It’s worth mentioning that the AA has another product too – the “RoutePlanner”. Although the route planner is “free” is has a value to the organisation both indirectly (by drawing traffic to the site that you might then cross-sell too) and intangibly (by offering a valuable service for “free” it enhances the brand).
Secondly, you need to identify the “product owners” for each of the core products that use your website as a channel to market, so in our AA example you’d probably have separate product owners for Membership, Insurance, Finance etc.
If you’re ever in doubt about who is, or isn’t, the right product owner then there is a simple test – “Do you have a significantly financial incentive (bonus) for Product “X” to succeed in the market?”.
The “Product Owner” Test:
“Do you have a significantly financial incentive (bonus) for Product “X” to succeed in the market?”
If the answer is “no” keep going up the hierarchy until you find someone who says “yes”. A product owner who doesn’t have any “skin in the game” regarding the success of their product line is a bad idea!
Thirdly, you need to work with the product owner to map out the product lifecycle of that product, and then identify the IT dependencies and deliverables at every stage along that product lifecycle. Within your product lifecycle you might also want to map out the “feature roadmap” if your product devolves into multiple features that will be released over time. Creating the “big picture” can be vital in motivating your teams and helping them to understand what you’re trying to achieve, and this in turn helps that to make better decisions.
Fourthly, you need to “sense check” your product lifecycle and feature roadmap against your business strategy and organisational goals. If they don’t align you either need to re-work the plan or you might decide to drop the product altogether and re-deploy those resources to a product that *IS* part of your core strategy.
Lastly you need to re-organise your DevOps teams around these products and align your delivery pipeline and processes with the product lifecycle (and feature roadmap). Your DevOps teams are responsible for the “inception to retirement” management of that product (*Top Tip – just like your “product owner” it might be a great idea to incentivise your DevOps teams with some “product success” (business) metrics in their bonus, not just technical metrics like “on-time delivery” or “system availability”. It never hurts for them to have some “skin in the game” to promote a sense of ownership in what they are delivering!).
So, to summarise, the key elements in a “product-centric approach” are:
(1) Breakdown your “website” into the key “Products” (that generate business value)
(2) Identify “Product Owners” with “skin in the game”
(3) Map out the product lifecycle (and ideally feature roadmap) for each product with them
(4) Sense check this product strategy with your organisational strategy
(5) Align your DevOps teams with the core products (and incentivise them accordingly!)
So next time you’re in a meeting and someone proposes a new “Project” see if you can challenge them to create a new “Product” instead!
image source: - CC - jeremybrooks via Flickr - http://www.flickr.com/photos/jeremybrooks/5401494223/sizes/s/