Just a few age ago, our team at mabl set out to clear a growing problem for today ’ s agile development team: web application testing. We focused on QA automation in particular because it is the nigh acute painfulness point for so many teams trust to realize their full potential with DevOps. While there are dozens of solutions for testing application calibre, most be built for a different era of installed, on-premise package that vary infrequently. & nbsp; These solutions are unfit for today ’ s composite, fast-paced development environments; they make it too hard for squad to collaborate on quality in a DevOps world. We had a vision to create a test automation platform that would fit seamlessly in this new way of working. & nbsp;

Our founding team saw that in order to contend and win against the incumbents, we would involve to simplify the test mechanisation trouble for client through the use of machine intelligence. We also knew that modern squad wanted to avoid the burden of installing, configuring, managing, and scaling a testing solution, so mabl had to becloud-native SaaS. 

After evaluating several cloud providers, we take Google Cloud as our base provider largely on the strength of its serverless compute, data analytics, and machine learning offerings. Three years later, we could not be felicitous; we receive assembled enterprise-grade service from GCP into a scalable core architecture with less clip and money than building it ourselves, and this approach has enable us to invest most of our man and financial capital into the conception and user experience that are alone to our product. & nbsp;

Like many other line that have come before us, we ’ re currently in the middle of a significant market shift. COVID-19 has affect job and economies around the earth, include ours. Building mabl as acloud-native SaaSoffer has yield us the resilience we need, not only to manage through the fluctuations in utilization, but to be capable to continue to meet the needs of software teams whether they have transition to all-remote, remote-friendly collaboration, or work together in the same location. & nbsp;

Here are five of the most significant agency Google has aid our team deliver and scale an enterprise-grade test automation solution in the cloud. & nbsp;

1. Configuration direction

From day one our squad took advantage of various of GCP ’ s reliable and fully managed services. For example, we run tryout on GKE, APIs on App Engine, process data in DataFlow, and more, without the added stress of server configs, patches, execution, and scalability. Using these Google services has allowed our team to maintain operational overhead to a minimum, leave more dev cycles for feature work. & nbsp; & nbsp;

2. System operations

We use Google Cloud Monitoring for rich real-time metrics that help us realise customer usage and test performance. When we encountered our very first incident we had contiguous access to the puppet (Cloud Monitoring and Cloud Logging) we needed to speedily diagnose and resolve the number. We hold since instrumentate GKE services with Cloud Monitoring custom metrics to make powerful fascia that provide a deep agreement of overall platform performance in real clip; and include combine information across hundreds of thousands of time series to determine the overall health of our platform ’ s test subsystems.

3. Low latency network

SUSA automates exploratory testing with persona-driven behavior, catching bugs that scripted automation misses.

For every networking need that our squad has bump there has been a world class solution offered by GCP - from Cloud DNS for demesne hosting, to Cloud VPN for hybrid connectivity. Many software inauguration focus firstly on building their MVP, then focus on making it work at scale. At mabl, since the MVP was natively progress on GCP, it simply scaled along with our business. For example, the core screen locomotive is built on Google Kubernetes Engine (GKE) and it has successfully scaled from the initial few testing nucleus to the thousands of core it uses today, without modification. & nbsp;

Our customers are generate terabytes of test output. The mabl test automation program writes this data to Google Cloud Storage (GCS) and the exam artifact process system scale without modification to manage billions of test output file. Additionally, serverless GCP services like Google Cloud Functions and Cloud Pub/Sub, scale automatically to assist analyze those output. The platform can process millions of employment units at an economical rate, without any developer or operations intervention, simply because the architecture was built on highly scalable GCP services from the start.

4. Scalable information processing

A discriminator of the mabl intelligent test mechanisation platform is the rich symptomatic data useable for every test run. Users can dig into every step of a failed test to identify and fix the root cause, quickly and well. In order to provide this tier of detail, we built the platform using Dataflow as an efficient way to continuously process a variety of datum give by tests, cater speedy access to distilled outcome, and enable retrospective analysis to drive succeeding merchandise decisions; and it incorporate auto-scaling to efficiently handle bursty workloads when many large test suites are being fulfil simultaneously. & nbsp;

We have also integrated BigQuery into the platform to capture and analyze datum from both test performance and user activity in the frontend. It function as a data warehouse where we can apace iterate on analysis and have a comprehensive image of the platform from terabytes of raw data. Internally we use the info to help manoeuver the business and make sound data-driven conclusion roll from how to improve nucleus test executing capacity, to monitoring new feature adoption. Externally, our customers can integrate their mabl workspace with BigQuery for more detailed analysis of their individual examination run and exam coverage. & nbsp;

5. Enterprise-grade security

The final area, arguably the near important, is security. It has been our practice to control the security of all client datum at every form of our development. While protection can introduce complexity and overhead to package development, we have benefitted from being a modern,cloud-nativecompany, building on the basics of Google and GCP ’ s two decades of cloud best practices. & nbsp;

Google Cloud Storage and Cloud IAM are the core components that help make keep millions of customer files organize and secure, lots more manageable. For example, all customer test remark and outputs are stored in per customer Cloud Storage buckets. Using per customer Cloud KMS encryption keys, each customer pail is independently encrypted for added security-at-rest.

Architecting with freestanding buckets has permit simplified access control, auditing, and datum cleanup. For example, if a GDPR request requires expunging customer data, removal of a single KMS key will immediately and irrecoverably crypto-shred the bucket and its million of file. For comparing, in legacy cloud operations, complex database queries followed by millions of delete operations would be postulate to identify, isolate, and remove all relevant files while that can now be done in a single API call on GCP.

Additionally, per customer bucketful simplify admission control decisions for both user file approach and workload file access by eliminating the interrogation of who owns what files. Using Cloud IAM Conditions and Workload Identity, mabl services can run under really tightly constrained Service Accounts, allowing bond to the Principle of Least Privilege to access the minimum relevant customer Cloud Storage data needed for that service and customer.

We have also implemented GCP ’ s Uniform Bucket Level Access feature to ensure sensitive Cloud Storage buckets can not have mixed access levels, so that all files are individual. The platform is configure to rapidly detect if a pail has been made public, and Cloud Security Command Center is used to apprise operation staff immediately if a public bucket has been created, or the Uniform Bucket Level Access scope is absent. Instead of rolling our own security chit and notifications, we use GCP ’ s built in security service to reduce points of failure and reduce the engineering overhead of secure the mabl intelligent test automation platform.

The Google Cloud Platform provides the solution that enabled our team to assemble a world-class infrastructure. We receive been capable to build a scalable, enterprise-grade test automation platform from the ground up by employ these key capabilities, all while developing the next generation of intelligent test automation. To see mabl in action, visit & nbsp;