Madrid DevOps – Continuous Integration

Madrid_DevOps2Madrid DevOps is a group of profesionnals dedicated to … DevOps, as you can imagine. There is a ‘Meetup Group’ where to find news, mainly about new meetings each month.

On April 10, the talk was titled ‘Continuous Integration’, with Manuel Recena Soto and Antonio Manuel Muñiz de ClinkerHQ. I asked some questions about their presentation, that you can find on https://speakerdeck.com/clinkerhq/integracion-continua (in spanish).

Hello Antonio, hello Manuel, may I ask you to introduce yourselves?

We are the leading developers of ClinkerHQ, an ecosystem of software development based on Open Source solutions (Jenkins, SonarQube, Redmine, SVN, Git, Nexus), ready to be used or installed as a service on a virtualized platform.

Many ClinkerHQ customers rely on us to establish or improve their practices of Continuous Integration (CI ) using our product. Our knowledge of CI comes from the experience gained during more than 7 years implementing, and helping to implement these practices in various companies.

What is Continuous Integration in your daily activity?

This is the heart of all our development cycle and if it stops beating, we can not continue working. Every day in Klicap, builds are performed, tests are executed, documentation is generated, snapshots are deployed, etc. and everything automatically. We work on a complex product where many separate software components are involved. Without a system of Continuous Integration doing that everything is available at once, it would be impossible to advance, and the day to day would become a maze of components and dependencies.

Continuous Integration notify us in almost real time when one of the pieces do not fit, and gives us the key to finding the cause of the failure instantly.

I see many people who think that just using Jenkins means they are DevOps, or that Continuous Integration is DevOps. What is your definition of DevOps ?

These are different concepts. Let’s say that CI brings to DevOps the same that it brings to software development, no more and no less.

For us, DevOps is a profile in a team, with one foot in development and other foot in infrastructure, so having the knowledge necessary to break that barrier that has historically always been between ‘development department’ and ‘systems’.

Is DevOps Continuous Integration + Continuous Deployment or is there more than that?

These are different concepts. Continuous Integration and Continuous Deployment practices are based on task automation to gain productivity. Can they be applied to DevOps? Of course.

I see people in favor of managing a dashboard with data about the quality of the infrastructure (availability, saturations, etc. ) for development teams. I agree with a few simple SLAs but I do not know if there is value beyond that (except to investigate an incident).

Monitoring the infrastructure is critical, especially when a product is offered as a service. In fact, one of the components of ClinkerHQ provides a series of graphs of time trends in resource consumption of the platform on which the product runs, this allows the user to make decisions about when upgrading.

There is a famous saying from Tom DeMarco that is widely spread in Quality Application: “You can not check what you can not measure”. Manuel, you say something like this in the presentation: “To get value, you need to measure”. Can you tell more about this?

Absolutely true. Here this phrase refers to software development tools and quality measurement that can be used today, as for example SonarQube. It is essential to measure how well (or poorly) you’re doing, and where are the weak points to improve. If we add the continuous measurement and troubleshooting through code inspection, then we have the practice known as Continuous Inspection.

What is the future of DevOps? What direction do you think or do you want that it goes?

The future of DevOps? Well no idea, but it seems that this is a profile that has come to stay, it has always been necessary. Perhaps automation tools and modeling insfraestructure (I am thinking of tools like Ansible, Chef, Puppet, etc.) have led to the creation of this model. These tools have made available to “those of the systems” the same opportunities that developers have for years now.

Thank you very much to both of you for taking the time to answer my questions.

This post is also available in Leer este articulo en castellano and Lire cet article en français.

Leave a Reply

Your email address will not be published. Required fields are marked *