How does continuous integration work in Salesforce DevOps?

How does continuous integration work in Salesforce DevOps?

Jason Mann on

Share with


Continuous integration (CI) is an essential aspect of DevOps. By implementing a highly automated Salesforce release process with CI, you’ll reduce manual effort and dramatically accelerate your release cycle. You’ll also make sure that your team is always in a position to release on demand — a key DevOps objective.

Given the benefits of automation, it’s no surprise that most Salesforce teams want to implement CI for their release pipeline. But success can be frustratingly elusive. In this post we’ll give a quick overview of the topic, but you can also download our complete guide to CI/CD for Salesforce:

What is continuous integration?

CI is the next step for teams that have adopted a source-driven process. If your team has built a release pipeline around Git for version control, you can introduce automation to speed up delivery and constantly check that the metadata in your Git repository can be deployed to Salesforce.

A CI workflow requires automation tools and a new way of working. Team members regularly merge their development work into version control, and the automation tool automatically deploys the latest metadata from Git to a Salesforce org. This automated deployment confirms that the metadata is deployable, and passes any tests the team set up on the CI job.

The challenges of continuous integration for Salesforce

For Salesforce teams with no experience of CI workflows, building an automated release pipeline can seem out of reach. Setting up CI using open-source automation software such as Jenkins takes time, effort and technical expertise.

Even Salesforce teams that are familiar with CI tools, thanks to experience of CI on other platforms, find it difficult to work around the quirks of Salesforce deployments. That’s because cross-platform CI solutions are great at automation, but offer no help when it comes to building a valid Salesforce package that can be successfully deployed.

In short, CI can seem like a mountain to climb with no clear route forward. Many teams don’t know how to begin, and many more are wrestling with open-source toolchains that take more effort to maintain than they save on deployments.

Salesforce teams want to solve continuous integration

The Salesforce ecosystem is moving decisively towards DevOps. The vast majority of Salesforce teams want to adopt source-driven, automated workflows for development and releases. Most have already adopted source control, and now they’re looking to make the transition to CI.

It’s vital that teams take those steps one at a time. Rushing to automate when a team is still getting to grips with version control rarely ends well. Maturing into DevOps is a proven approach, while the ‘big bang’ model for DevOps implementation almost always fails.

While our research found that 82% of Salesforce teams hope to have automated release pipelines by the end of the year, plenty of teams are stuck when it comes to CI. Despite having automation tools in hand, they haven’t yet found success with them. This is clear from the sheer amount of teams that routinely have to fix stalled CI jobs. Just 20% of teams say that they hardly ever need to fix a stalled CI job, while the other 80% have to do so as a matter of routine.

Stalled CI jobs are common for Salesforce teams

What causes CI jobs to stall?

Only repeatable, reliable processes can be successfully automated, so until your team has a solution to the problem of Salesforce deployments, which are notoriously error-prone, the pain of stalled CI jobs won’t go away.

Teams need to troubleshoot the problems that regularly cause their deployments to fail before attempting automation. Certain Salesforce metadata types can be tricky to deploy, and most deployment tools don’t handle those quirks automatically.

There are some changes that will cause a conflict in Salesforce and so won’t be deployable, but Git will happily merge them because they don’t touch the same files. Sometimes there are dependencies between different feature branches and merging only one of them will cause the CI job to fail. Validating that Salesforce can receive changes before deploying or merging them is one of the most effective ways to resolve these problems.

How CI jobs run smoothly in Gearset

Gearset’s continuous integration combines the power of automation with our deep understanding of Salesforce metadata. Gearset’s deployment engine has built-in ‘problem analyzers’ that find and fix common errors in deployment packages. The result? Effective automation that’s easy to set up and reliable enough to leave running without manual intervention.

Each time a CI job runs, Gearset automatically compares your environments, selects metadata based on the filter you’ve configured, and resolves any deployment-blocking errors. It then builds the package, runs your tests, and deploys to the target — or validates the package in the target. You choose when and how to be notified about the status of CI job runs.

You can assemble and manage your automated release pipeline for Salesforce using the pipelines view within Gearset.

Pipelines view of continuous integration in Gearset

Continuous integration produces results for Salesforce teams

It can be tempting to stop short of implementing an automated, source-driven release process for Salesforce. Taking the pain out of deployments by graduating from change sets to an intelligent deployment tool is such an easy win. Adopting source control and CI is harder, as it requires learning new concepts and cultural change. But successfully automating a source-driven release process unlocks the most valuable benefits of DevOps.

Once teams have reliable automation in place, work items flow along the release pipeline unhindered, and different workstreams are continuously integrated without the constant pain of merge conflicts.

The benefits that teams want to see from DevOps mostly materialize after introducing automation. Of the teams that hardly ever struggle with their CI jobs, 91% report larger Salesforce ROI thanks to DevOps and 70% fix bugs within a day. And when it comes to achieving the benefits of DevOps, teams using tools for continuous integration and continuous delivery (CI/CD) outperform everyone else in every category.

Salesforce teams with CI/CD outperform all others

Experience CI success first-hand!

You can try out CI — along with everything else in the app — as part of your free trial of Gearset. Still wondering why 82% of Salesforce teams are working towards implementing continuous integration? Download our CI/CD for Salesforce ebook to get up to speed!

Try all of Gearset for free