Latest Posts

Announcing Kasten K10 Version 2.5

Kasten v2.5 - Whats New

The need for contingency planning has never been more visible! It is exceedingly important to enable application portability so that your operations remain unhindered. With accelerating production deployment of Kubernetes applications, we are excited to announce the availability of version 2.5 of K10, Kasten’s cloud-native data management platform. Some of the world’s largest enterprises are already using K10. The new version adds a unique Cloud-Native Transformation Framework with two very significant capabilities for backup and mobility for Kubernetes applications: Data Transformations and Application Transformations. These capabilities build on the security and operational simplicity pillars of K10 to bring about cloud-native scale for data and deployment portability and flexibility.

Data Transformation Engine

Applications have benefited from the shift to cloud-native microservices and much has been written about this topic. However, we also wanted to bring about the same benefits to the underlying data transformation engine that includes critical data management functions such as data movement, dedup, compression, etc.

Application-centricity has been at the heart of Kasten’s design right from the start. We have now taken this to the next level where K10 v2.5 gives every application its own data transformation engine. As a result, this cloud-native Data Transformation Engine is faster, smaller, and safer. The figure below shows some of the key principles we employed while building this and how that compares to legacy approaches designed as monoliths for hypervisor environments.

ArchPriciples-KastenDataEngine-White

Let’s take the critical task of deduplication as an example.  With this novel micro-services based approach, the Data Transformation Engine realizes the following benefits:

  1. Increased capture efficiency: with each application instantiating its own engine, tasks can be parallelized with parallel transfers, optimized hashing, and smaller fault domains.
  2. Serverless, source-side dedup: allows scale-up but also scale down to zero so there are no idle cycles. No need to dimension for peak capacity that legacy client-side implementations were designed for.

Application Transformation Engine

Kubernetes has brought the promise and benefits of application portability to the cloud-native ecosystem. Portability is an overloaded term and, as an example, could mean application portability across:

  •  Storage systems and Kubernetes versions while restoring applications
  • Kubernetes namespaces and Availability Zones (AZs) while cloning applications
  • Regions and Cloud Providers while migrating applications

To truly enable application portability across critical data management use cases including restore, clone and migrate we need to handle transformations across:

  1.   Infrastructure
    • e.g., Storage type change on restore or clone (e.g., spinning disk to SSD)  
  2.  Kubernetes Environments
    •  e.g., Production to test environment clones where DNS names within the application need to change
  3.  Application Specifications:
    • e.g., Changing portions of the application’s schema to remove stale secrets or update the specs to reflect schema definition updates

This gives enterprises the freedom of choice across various infrastructure, clouds, and Kubernetes distributions. The Application Transformation Engine empowers the CloudOps teams to provide these application portability benefits at a massive scale. Automation by leveraging rich APIs and policies exposed by K10 allows for an extremely efficient DevOps approach.

The figure below shows a sample Kubernetes application on the left. Under the hood, the application is composed of microservices that are operating on various Pods. Some of these Pods, in turn, are associated with v­­­arious storage volumes via PV-PVC bindings, while other Pods are running databases such as PostgreSQL. Additionally, there are usually hundreds of instances of other Kubernetes objects including ConfigMaps, Secrets, etc. that contain application constructs that may need to be transformed. K10 treats this entire Kubernetes application as the unit of atomicity when applying transforms and provides portability benefits across infrastructure and distributions.  

K10-TransformEcosystemSupportWithApp

You can read more about the transform operations (including Test, Add, Copy, Move, Replace) and realize their power by referring to the K10 transform docs here. We have also extended the Kasten K10 dashboard GUI to surface these powerful transform constructs with support for advanced resource selection capabilities such as regular expressions and all the transform operations mentioned above.

K10DashboardClip

Bringing it together

 

Let’s take a real customer example to surface the benefits of the Kasten K10 Data and Application Transformation Engines. In this case the customer’s CloudOps team wanted to migrate more than 170 applications (composed of more than 18,000 components) running on RedHat OpenShift to another cluster on AWS. In addition to a significant change in Kubernetes distributions (OCP v3.11 to OCP v4.3 across 5 Kubernetes releases), container runtimes (docker to cri-o), and container registries, the underlying storage also changed from a mix of EBS and an external Ceph cluster to EBS and a Kubernetes-native Ceph cluster administered via Rook.

CustomerMigrationExample

Since the migration needed to be done with minimal disruption to the various applications owners the actual migration was done over a weekend! This scale and limited time needed the underlying cloud-native scale-up and also the scale-down-to zero capabilities that we alluded to in the design principles of the Data Transformation Engine.  

The varying infrastructure and Kubernetes environments also required a variety of application transforms. The table above enumerates some of these application transforms applied to various resources via automation policies for this migration. The K10 Application Transform Engine enabled just two ops personnel to successfully complete this migration during the weekend and support 700 developers without the need for professional services or application changes!

Come join SopraSteria and RedHat as we present this here.

The capabilities illustrated using the migration use case above are as applicable for other data management use cases such as Restore and Clone across storage systems, namespaces, AZs, clusters, cloud providers, etc.

 You can try the fully-featured and free Starter Edition of K10 today. We are also available in the EKS and the GKE marketplaces (free trial option available). To learn more about K10, please visit the product page, check out the K10 datasheet, watch our recent webinars, and read the documentation. Finally, we would love to hear from you to see how K10 could be useful in your Kubernetes environment. Find us on Twitter, drop us an email, or swing by our website!

Gaurav Rishi

Gaurav Rishi is Head of Product at Kasten. He previously led Strategy and Product Management for Cisco's Cloud Media Processing business. In addition to launching multiple products and growing them to >$100M in revenues, he was also instrumental in several M&A transactions. An engineer at heart, Gaurav is a computer science graduate and has an MBA from the Wharton School.


Share

Recent Blog Posts