CdCon and Spinnaker Summit 2021 Summary
This blog will recap the current trends in continuous delivery showcased at cdCon and Spinnaker Summit 2021. Secondly, we will give a glimpse of some customers who have knocked off challenges in their SDLC by using Spinnaker.
Interesting trends from the cdCon event to note are:
- There Still a Need for Speed: Many sessions highlighted that a sizeable amount of organizations are still struggling to achieve velocity in moving code to production in their software delivery. And the reason can be mandatory manual checks or security checks or lack of best practices, or simply the combination of all.
- Before Trust, Do Verification: There were few sessions that highlighted the power of automated verification- a faster and accurate way of identifying risks of software. These sessions also described how to use machine learning for automated canary analysis- a vital subject for organizations at the forefront of their continuous delivery initiatives.
- Being better than Yesterday: Another common concern that emerged among organizations, who have scaled their CD implementation, is how to enhance interoperability among CI/CD tools to make security and audit a part of the SDLC. Every organization wants to check for vulnerabilities without interrupting the speed of delivery. And everybody wants finger-tip information to trace and find out the cause of a bad deployment.
- GitOps- The new guy has a chance: With the rise of usage of Kubernetes, small and non-regulated organizations like the ability to maintain the intended state of applications in version control like Git and automatically deploy into Kubernetes project from Git. GitOps style is excellent for small and simple deployments, but for complex deployments, progressive delivery, or continuous delivery, there would additional requirements of interoperability with other tools.
Watch the recap of cdCon Spinnaker Summit 2021 by Balaji S, VP Products, and Gopinath Rebala, CTO, OpsMx.
We have all the exciting sessions from cdCon in case you have missed them.
Automated Decisions to Achieve Speed and Stability with Continous Delivery– Gopinath Rebala, CTO, OpsMx
Decisions are made at different points and by different actors. For example, developers want to move code to integration env, testers want to move artifacts to test environment, and the Ops team wants to move the application to production. And they all rely on data to make the decision. However, the variety of formats and sizes of data and many DevOps tools emitting those data make decisions complex and time-consuming.
Gopi discusses how machine learning can be used in continuous delivery to expedite the decision-making process by consolidating data and providing risk of application in each stage.
How AI/ML is Making a Difference in CI/CD Domain– Anjali Batra, Principal Data Scientist, OpsMx
Manual judgment is not enough when a vast amount of data must be understood and critical decisions must be made. For example, the Ops team has to analyze torrents of metrics and logs and decide if a canary is fit for production release.
Anjali Batra, Principal data scientist, outlines various AI algorithms for logs and metrics for risk assessment and seamless canary analysis. You will learn the following items:
- What is supervised (classification) and unsupervised learning (clustering) and how to leverage these algorithms to analyse logs for accurately predicting the risk of a release.
- How to score canary wrt to the baseline to decide rollback or roll-forward
- Various weight modeling and analysis techniques for metrics to find real-health of applications
- Contextual correlation between logs-to-logs, metrics-to-logs, and metrics-to-metrics, to diagnose and find the root cause of regression.
Spinnaker @Cisco- Scaling to Support Large Diverse Applications
If a large-scale successful continuous delivery project has to be referred to, then Cisco’s success story is remarkable with the following results.
- # of Spinnaker pipelines used for multi-cloud deployment: 4000+
- # of microservice managed: 1300+
- # of deployments/year: 68,000+
- Target Environment: SaaS (Salesforce, Oracle, Workday, ServiceNow, Snowflake), major public clouds, on-prem VMs and OpenShift clusters
- # of developers using the pipelines: 1000+
Besides faster application deployment, Cisco leverages OES and Terraform to spin up Kubernetes clusters and EC2 instances on-demand across dev and test environments, further speeding up their software delivery.
The noteworthy point is, while Cisco has attained break-neck speed, they have ensured the software delivery is secured, compliant, and out of risk.
Spinnaker Success Stories: Interswitch and a Large Telecom Company
Interswitch and a Large Telecom Company share their story on how Spinnaker has emerged as a leading tool for achieving scalable multi-cloud deployments. Spinnaker allows them to enable self-service, control (through policy integration), and deploy safely with automated verification.
“Payment company has many financial regulations and audits, so we wanted a single CD tool to tackle our policy requirements and handle multiple deployments per day. We considered Spinnaker because it is open-source and also flexible. “
– AbdulBasit Kabir, DevOps Engineer, Interswitch
“Other tools just do kubectl -apply, but Spinnaker provided us visibility like post-deployment status information, and safe deployment strategies and ability to find out risk while deploying into OpenShift clusters.”
– Marc Andre Vanier, Business Solution Architect, Large Telecom Company.
Watch the below video to learn how Spinnaker helped Interswitch and Large Telecom achieve radical transformation in software delivery in scalability, velocity, quality, and safety.
Spinnaker Workshop: Source to Production with Rollout Strategies
If you deploy to Kubernetes, AWS, GCP, or any on-prem VMs, we have an exciting and informative workshop with best practices to help you with your Continuous Delivery journey.
Watch the Spinnaker workshop to learn how to orchestrate software delivery- from the code commit in Git, build process in Jenkins to deployment in Kubernetes, AWS, GCP, and Azure. And also learn how to apply safe deployment strategies like canary and blue-green for progressive delivery.
OpsMx is a leading provider of Continuous Delivery solutions that help enterprises safely deliver software at scale and without any human intervention. We help engineering teams take the risk and manual effort out of releasing innovations at the speed of modern business. For additional information, contact us.
