Exclusive: Strengths and limitations of the AWS/Cisco partnership
Continuing the trend of data center giants forming partnerships, earlier this month AWS and Cisco revealed they had teamed up to offer a new hybrid cloud platform.
Cisco is now offering a hybrid system to run Kubernetes on AWS to provide an integrated platform that the companies say will simplify the process of building, securing, and connecting Kubernetes clusters across private data centers and the AWS cloud.
It seems like a match made in heaven for businesses struggling to digitise while juggling between cloud and traditional data centers. However, Iguazio CTO and co-founder Yaron Haviv says while there are certainly some positives from the partnership as hybrid solutions gain traction, there are also some limitations.
"The shift towards hybrid solutions, which provide high connectivity between cloud and edge, is important, as customers want to leverage the cloud for elastic compute while still running time- and security-sensitive projects at the edge," says Haviv.
"We also see that Kubernetes is becoming the de-facto infrastructure API and believe that it is the right layer for abstraction to enable a multi-cloud and hybrid cloud infrastructure. However, Cisco should think more about the higher level abstractions which enable faster and easier digital transformation.
Haviv says in the cloud, companies are moving to data services, AI, and serverless functions used directly by developers.
"However, there are no equivalent abstractions in Cisco's offering which instead focuses on technology building blocks and perpetuates endless IT tasks. This forces IT orgs to build and integrate a high level service layer from an endless number of new technologies and open source projects, and in most cases they don't have the skills or resources to do so," says Haviv.
"Cisco provides a more modern infrastructure, but not real cloud-like services, while users are looking for higher level, cloud-like abstractions which include managed data services, serverless functions, AI and data-science tools. Those tools need to be integrated into a homogeneous solution with end-to-end security just like in the cloud.
Haviv is confident in the future of hybrid and edge solutions for businesses of varying sizes and needs.
"Companies with real-time and fine-grained security needs must process data locally to overcome the cloud's latency and bandwidth constraints. We believe that instead of on-premises deployments, hybrid and edge solution will dominate and the edge will be an extension of the cloud," says Haviv.
"Thus companies will be able to consume data services, AI and serverless at the cloud and at the edge, leveraging elasticity in the cloud and high performance at the edge.
Of course, as with any technology there are still obstacles to be overcome, and Haviv has identified his view of the most prominent.
"Vendor lock-ins lead to a lack of standard abstractions across cloud, edge and different products," says Haviv.
"Additional challenges include data and workload mobility across cloud and edge with low latency (which serverless is now solving), processing large data volumes at the edge and providing intelligence at the edge so that this process is done seamlessly with minimum IT hassles.
As aforementioned, a growing trend among the big players in the data center and cloud industries is to partner up in their solutions, which would seemingly be of detriment to the smaller providers. Haviv however, says this isn't the case.
"Smaller companies have the advantage of no bureaucracy which lets them move fast and think out of the box. Even Cisco has more than once invested in new companies to speed up time to market of new technologies," Haviv concludes.
