AWS Developer Tools Customers

It was easy to adopt AWS CodeArtifact. In a couple of hours, we scripted the creation of the domain and repositories and also changed a couple of IAM roles to add access. With just that, our team and our AWS CodePipeline continuous integration system had access. It also helped that CodeArtifact supports the primary package managers we use. Because access is managed through IAM, there were no additional logins to juggle and no secrets to manage in the pipeline. Just as important is that CodeArtifact is an AWS managed service. There is nothing beyond our own code to manage, no patches, no servers, no networking, and no subscriptions.

Our platform uses numerous Node.js AWS Lambda functions that have dependencies on open-source packages. We find it hard to quickly resolve issues in third-party packages because of our limited control. For example, a defect in one of the open-source packages for connection pooling and management caused our application to have intermittent database connection failures. We didn’t have time to wait for the upstream package author to fix it.

AWS CodeArtifact, as a transparent proxy to the upstream npmjs.org, enabled us to patch the package locally and upload it to our CodeArtifact repository. The beauty of this approach is that it resolved the issue without us needing to update the source of any of our numerous Lambda functions - a huge time saver for serverless and microservice scenarios.

Previously, our test runner was setup on Jenkins. Our test suite was large enough that we had to parallelize the running of tests to run it in reasonable amount of time, which is around 15 minutes. However, if more than one developer checked-in their code, they would have to wait for their turn to run their tests. On busy days, the wait times would reach as much as 4 hours! This process was highly inefficient. With AWS CodeBuild, we can run as many parallel streams of tests as we want. There is no need for any queue or developers to wait for hours. CodeBuild elastically scales the number of containers as necessary.

We chose to use CodeBuild as it is a scalable build platform that eliminates the complexities involved in provisioning and management of CI infrastructure, without needing any customization. CodeBuild’s functionality is simple to understand and met our needs precisely. It easily integrated with Jenkins and allowed us to run replicable builds in an asynchronous manner. With CodeBuild, we could run multiple builds at the same time, eliminating all pipeline bottle-necks in our CI/CD process. Additionally, it allowed us to specify custom environment parameters on a per build basis and thus be more confident about the integrity of each build.

We had problems with scalability and concurrency. We had to wait in long queues to run builds, often around an hour. We needed much more machine resources, but since we only run builds during the day, we wanted something that provides only the resources we need when we need them. CodeBuild lets us fully customize our build tasks, automatically runs builds concurrently, took the pain out of managing the build environments, and we only pay for what we use. In the end, CodeBuild helped us decrease build times dramatically.

We expect to save $32 million in the first five years by running our website and other critical applications on AWS. For a growing company like ours, it is incredible that we are able to bend the infrastructure cost curve down as we continue to expand.

We’re growing at 300% annually, and we have the ability to meet that growth using AWS. We can scale our service as much as we need to and as fast as we need to.

SmartNews uses machine learning to evaluate tens of millions of articles, social signals, and subtle factors in real time to uncover which stories are worth reading in any given location.

AWS CodeDeploy helps us ship our code to customers faster and more reliably.

We are fully supporting our continuous integration and deployment pipelines by relying on AWS, and that has really boosted our development efficiency. Using AWS, we’ve gone from deployments taking six weeks to one per week, and very soon we expect that to be multiple deployments per day.

When we detect anomalies in production, we have been able to reduce the investigation time from days to hours and sometimes minutes thanks to AWS CodeGuru’s continuous profiling feature. Our developers now focus more of their energy on delivering differentiated capabilities and less time investigating problems in our production environment.

Our use of CodeGuru Profiler is very simple but it has been extremely valuable to us. We've optimized our worst performing service to reduce its latency by 15% to the 95th percentile in a typical business day.

Any continuous integration and deployment pipeline should be automated, easy to manage, and discoverable, and that’s exactly what we get using AWS. We get a level of simplicity and transparency we simply couldn’t have in our previous on-premises environment.

We chose AWS because of its leading position in the market, its reliability, and its pay-as-you-go model.

Now our engineers can consume pre-approved AWS architecture that are feature rich, including AWS Fargate services that include fine-grained deploy roles, load balancing, certificate validation, autoscaling, AWS Lambda functions to forward logs and a choice of Dynamo or RDS as a storage backend. We can now feasibly do in hours what used to take us up to six weeks.

CDK has been a gamechanger for us. It has drastically improved our feedback cycle and reduced the time it takes to go from brand-new to fully-deployed infrastructure.

By moving to AWS, we no longer have to worry about cache flushes or capacity issues. We have the scalability and performance we need to deliver high-quality real estate images, which is so important to the Zillow user experience.

Using AWS Elastic Beanstalk, the transformation to AWS was incredibly easy. I was sure it would be much more difficult to migrate all our servers, but we went from being in production on our old provider to being in production on AWS in just two weeks.

With AWS, it’s now possible for us to create new products and attract new users without having to think twice about the consequences for our IT infrastructure.