Overview — Working Together
There is a prevailing myth that DevOps and IT Service Management (ITSM) and the IT Infrastructure Library (ITIL) are incompatible. However, this supposition has very little basis. ITIL is a framework from which you can take or leave shares you like and, in fact, this framework delivers many useful models for DevOps implementations.
There’s essentially a lot of synergy between ITIL and DevOps. If we understand ITIL as a procedure framework and see DevOps as, mainly, a culture of collaboration, there is no motive we cannot have a process framework participate very well with a culture of collaboration.
Many others have noted the interactions between DevOps and ITIL. On the Atlassian website, Barclay Rae writes, “We need the key elements that are found in both ITSM and DevOps, whether we use these explicitly or not. DevOps is much more than just automated development; it involves collaboration and a blame-free culture. As well, ITSM/ITIL shouldn’t be considered as an administrative burden, but rather used in an agile way. ITIL, in specific, isn’t perfect and needs a more modern finish, but the core performs are sound and proven.”
DevOps and ITIL Help Companies Scale
The size of the organization does problem when applying DevOps method to ITIL. When an initiative is moving faster, they actually need to have the process framework in place, else, you’re going to get lost. In a 5 or 10 person shop, it is much easier for everyone to know what everyone else is doing at any given time. With smaller companies, there is less need to train everything in details. As companies grow it becomes imperative that they have the processes to track what’s going on and react when problems arise.
Suggested Read: Keys to DevOps success
When shops are doing 15 to 30 releases a day they want to have a record of releases, so when approximately breaks, they can rapidly and easily go back. For mid to large sized companies we need to know what’s been changing, what the process is, who needs to get involved so that we’re engaging the right people but not all the people, and we’re replying quickly to get services restored. Advances in operational telemetry supported by DevOps evangelists goes elsewhere this to capture all actions, with deployments, auto-scaling actions, etc. We make capturing all of those parts of the data flow, so there is a lot of data.
Incident management is an alternative great example of the importance of process frameworks at scale. When you’ve got millions of moving pieces, when something breaks, it is significantly more important that you have a framework for responding appropriately to that incident. Because at small scale, everyone knows who wrote the code – it’s may even be the same engineer that set up the server on which the code runs. At a large scale, we must have a procedure in place to coordinate across different systems, different teams and different geographies.
Problem management becomes critically important too. Because to be good at collaboration and DevOps engineers are also constantly doing some introspection and looking at the underlying causes of problems to learn from them. The best DevOps shops are the ones that are not just reacting when a problem occurs, but actually being proactive and making sure problems don’t occur in the first place.
The ITIL method is just a framework. It is not a command that ‘thou must fill out twenty pages of documents to statement software’ although, some very bureaucratic implementations definitely make it finish that way. When we know the IT Infrastructure Library as a framework it becomes apparent that there is no purpose we cannot relate DevOps principles in this framework to make our technology operations even better. By enchanting a DevOps approach to our development definition and implementation we can effort safer statements quickly, we can confirm better communication between our teams, drive quicker determination to incidents, and we can ensure we keep our focus keenly on delivering value to our customers quickly.