How to Spur Line of Business Adoption of Kubernetes
The key to line of business developer adoption of containers will be to provide self-service for Kubernetes CaaS (Container as a Service) platforms. However, vanilla Kubernetes by itself will not suffice to drive ease of use and developer experience. Developers must be able to easily deploy runtimes as shown below and also manage their lifecycle.
These runtimes will be used for three common reasons –
-
Add application functionality using runtimes such as RDBMS, NoSQL, BI visualization tools, Message Queues, etc. This includes MySQL, MongoDB, Tableau, Kafka, etc
-
Higher-level microservices-based functionality using Service Mesh, CI/CD tools, etc. This includes platforms such as Istio, Spinnaker, JenkinsX, etc
-
Operational use and visibility (monitoring, logging, audit, etc) to enable DevOps teams to easily deploy and manage the lifecycle of additional applications and tools for monitoring, tracing, logging, etc. This includes platforms such as Prometheus, Jaeger, etc
With Kubernetes gaining a reputation for being hard to operate at scale, it is important to obviate the need for admin and SRE teams have to learn every new integrated app shown above from scratch (understanding API’s, performing configuration management using complex YAML and Helm charts, failover/ HA processes, setting up persistent storage, backup plans, and more) – so they can manage these on an ongoing basis for mission-critical environments. It is key to ensure the smooth operationalization of such services but also the seamless management by operation teams.
Enter Managed Runtimes by Platform9. Built on upstream open source technology, Platform9 is leveraging the capabilities in the Kubernetes community, including Kubernetes Operators. These are extended, for various managed applications, to allow organizations to benefit from a fully managed service and SLA, without having to deal with the overhead and complexity of managing multiple applications, operators, and their inter-dependencies for production deployments.[1]
Introducing Kubernetes Managed Apps
The goal is to ensure that developers can now use Prometheus, EFK, MySQL, and other apps – delivered as a service, with 99.9% SLA on any infrastructure (on-prem, or in the public/hybrid cloud).[1]
This enables DevOps teams to provide self-service consumption of these applications, simplify Day-2 operations, and ensure end-to-end SLA for both the Kubernetes infrastructure AND critical applications and tools.
Per customer demand, the first set of fully-managed applications to be offered are:
-
Managed Prometheus: Platform9 Managed Prometheus provides Dev and Ops teams with easy deployment and configuration of the Prometheus stack and the Alert Manager as a multi-tenant service. The service is HA-ready, with all persistent storage, automatic backups and recovery handled automatically, with no need for Admin intervention.
-
Managed EFK: The first introduced Fluentd would soon be followed by Elastic and Kibana for an integrated managed EFK service for logging and indexing.
-
Managed MySQL: The popular relational database for stateful Kubernetes applications would be delivered using Percona’s MySQL offering.
Additional applications will be added on an ongoing basis in future releases.
Managed Apps Catalog
A Powerful Enterprise-grade User Experience
Each managed app provides one-click deployment for developers while giving Operations teams the governance they require to manage security, access controls, and capacity across the entire organization.
-
Self-service: Users will be able to choose from a variety of applications available in the Kubernetes Managed Apps Catalog.
-
Batteries included: Applications deployed on Kubernetes typically requires integration with Prometheus for monitoring, EFK for logging, persistent storage for stateful deployments, etc. Managed Apps offering provides these core services as built-in capabilities, provided out of the box for each deployed managed application. This way, self-service users do not need to worry about installing and configuring these dependencies while deploying their specific app instance. In addition, monitoring and logging of each managed app are enhanced to incorporate system-level metrics around resource utilization of the app, with no involvement or configuration required from end-users.
-
Integrated, end-to-end SLA: This covers the lifecycle management of the managed application instances – including version compatibility and upgrades across all instances – as well as all the dependent components required to run the managed app. This includes 24/7 proactive monitoring around application uptime, auto-detection and correction of failures that the app instance might run into, security patching, performance guarantee, support by Platform9 Kubernetes Certified Engineers and Application experts, and more.
-
Support for multi-tenant, multi-cluster or dedicated instances deployments: IT Ops can easily deploy and configure any managed application available in the catalog in a multi-tenant, multi-cluster fashion, or as dedicated instances per each cluster. For example, they can enable Prometheus monitoring on all clusters, with a click of a button. That way, as more applications are developed, they are enabled with monitoring out of the box, with no additional overhead.
-
Full RBAC capability for individual namespaces or cluster-wide enables IT Ops to easily let developers access and self-service their own managed applications.
-
Simple UI or Kubectl CLI: All managed applications can be deployed and administered either through a simple UI or with the native Kubectl CLI interface – instead of having to deal with separate, complex YAML or other configuration files/tools for each service.
-
Operations teams benefit from centralized visibility, management, and governance across all Kubernetes services – such as monitoring, logging, storage, service mesh, and more.
Conclusion
The new Managed Apps offering enables enterprises to easily adopt both Kubernetes–and Kubernetes applications–at scale, with no management overhead for the underlying infrastructure stack or its related services and tools. It is designed to strategically tackle the Kubernetes complexity challenge by extending Platform9’s seamless management experience of Day-2 operations to the applications themselves, eliminating the need for DevOps teams to become experts in these related tools and services.
References
[1] Kubernetes Managed Apps: Prometheus, EFK Stack, MySQL, and More – Delivered as a Service, with 99.9% SLA – https://platform9.com/blog/kubernetes-managed-apps-prometheus-efk-mysql-more-delivered-as-service-sla/