Data protection continues to gain more attention from the investment community. In the past few weeks, we’ve seen Clumio come out of stealth and raise $51 million for cloud-native data protection. Kasten has raised $14 million for its application-focused backup solution using Kubernetes containers.
Clumio emerged from stealth on 13th August 2019, bringing to market a data protection-as-a-service solution (DPaaS). The company was founded in 2017 by three former PernixData employees, including PernixData co-founder Poojan Kumar. PernixData was acquired by Nutanix for an undisclosed sum in August 2016.
In the first instance, Clumio will offer protection for virtual machines, running either on-premises (in VMware vSphere) or on VMC (VMware Cloud for AWS). In both cases, the Clumio solution is taking advantage of the vSphere data protection API using a proxy on the client-side as a data mover. The proxy also de-duplicates, compresses and encrypts data before transmission across the network to the public cloud. Charging is based on the number of virtual machines protected by the user.
Kasten is a data protection solution for Kubernetes environments. The company was founded by Niraj Tolia (CEO) and Vaibhav Kamra (VP engineering) in early 2017. Kasten recently announced a $14 million Series A funding round, with a total of $17 million invested to date.
The K10 platform is designed to back up Kubernetes environments and itself runs on the Kubernetes platform. This includes data movers and the management GUI. K10 can also be used for application migration, copy management and disaster recovery. One interesting feature is the use of Kanisters, definitions that provide the capability to implement application-specific backups. Kanister blueprints are created within an open-source framework and so can be extended and developed by the entire Kubernetes community.
These new solutions highlight both a trend and a challenge in modern data protection. The landscape for application deployment continues to widen, with new frameworks like Kubernetes and new locations like the Edge. IT organisations don’t want to install data protection software and hardware in every location the business operates. This makes SaaS-based solutions great for dispersed organisations within increasing amounts of edge computing.
However, container-based solutions are gaining wider adoption and data protection at this level needs to be focused on the framework itself. We need to be able to backup data and application definitions, rebuilding them as necessary, either for recovery or simply to move workloads around.
In briefly reviewing these two solutions, we see an immediate challenge. Modern application deployment offers us a choice of platform, solution and vendor. In many cases, these solutions are meshed together, with some data on physical servers, some in virtual instances and some in containers. Equally, these solutions could run on-premises or public cloud (or both).
Somehow, we need to ensure that we don’t introduce technical debt, stranding protected data in only one place or another. Active data typically resides on a single platform at any one time. Secondary data lives over months and years, so more thought is needed as we look at long-term retention.
The Architect’s View
Clumio is entering a market where SaaS-based backup already exists. Druva has set the benchmark here, although there are a plethora of smaller companies offering SMB/SME protection too. Vendors such as Rubrik and Cohesity already exist in the public cloud and are spanning on-premises and cloud locations. They are also offering value-add services for historical data stored on their platforms.
Convincing enterprises to move away from on-premises solutions like Commvault or Netbackup won’t be easy. There’s a lot invested in existing backup solutions and even if a customer did move elsewhere, there’s a problem with managing historical restores that must be addressed. Clumio will need to use its $51 million war chest to bring new features to the platform as soon as possible. This will be a significant challenge to achieve.
Kasten has a novel approach that has wider applicability to data and application mobility. The question here is how portable application data will be. If the Kanister idea can gain widespread adoption, then the solution could be very interesting indeed, especially if it can be made applicable to other backup solutions that aren’t just based on containers.
I’ll be reviewing both solutions for inclusion in the Architecting IT white paper “Modern Data Protection”, version 1.0 of which will be released on 1st October 2019. You can register to receive this report once it’s produced by following this link.
Post #36b7. Copyright (c) 2019 Brookend Ltd. No reproduction in whole or part without permission.