Occasionally, I run into an interesting company at the beginning of a critical trend. When I got a call from Tetrate PR that described this young firm, which was initially funded by companies like Dell and Intel, it reminded me of VMware and hypervisors. Hypervisors were initially created to deal with software problems that arose when Intel iterated its platform. The initial hypervisors allowed existing software to run on new Intel hardware without being rewritten. This relationship between hypervisors and hardware made Pat Gelsinger’s move from Intel to VMware and then back to Intel ironic.
A similar problem has emerged in the multicloud world, but it has more to do with needing software to run on multiple cloud platforms to assure redundancy and uptime. As we have seen, almost any cloud provider can have a catastrophic outage. Maintaining on-premises resources to handle such an outage is wicked expensive; having another cloud provider on hot standby is far cheaper. But each cloud provider is different, which means that applications that run well on one cloud provider won’t run well on another unless altered or rewritten. What Tetrate has, and likely why both Intel and Dell invested in them, is like a hypervisor for the cloud, allowing applications to move freely, secured, and effectively managed across cloud providers. Call it an entirely different kind of hybrid HPC computing solution where rather than shifting between on-premise and the cloud, you shift between cloud providers as needed.
The Idea Came From the Cloud
It is fascinating to note that this solution, which Tetrate calls a “Service Mesh”, grew out of what cloud providers have to do themselves. Often, they use various hardware (some they may specify themselves) and providers to both optimize their environment for their customers and benefit from competitive bidding for the related servers. They employ a control, management, security, and compatibility layer that allows them to move customer workloads to different sites, different hardware, and even different parts of the world as needed.
For lack of a better term, middleware or cloud hypervisor, is typically proprietary to the cloud provider because it equates to either a competitive advantage or table stakes for the service. There is no real interest in licensing it out because people would use it to either migrate to a competitor as needed, or bring in lower-cost providers reducing the potential revenue for the cloud provider that shared the technology.
It isn’t that providers don’t recognize a need. Whenever there is a significant cloud outage, that need is self-evident, but the financial risk of sharing the technology exceeds the individual benefit. Granted, this is a legacy that came before open source rose in popularity — called lock-in — but it still exists in executives’ minds today.
Tetrate’s Service Mesh
Tetrate took this portability concept and turned it into a subscription offering that provides traffic control, assures reliability and uptime by using redundant cloud providers, reduces latency and error rates in cloud operations, and secures via pervasive encryption the result.
One of the significant benefits is that developers in a Tetrate shop don’t have to worry about these elements as they are handled and managed by the Tetrate layer. This result is developers working with a hypervisor don’t have to worry about hardware inconsistencies because that hypervisor handles those. This outcome helps with operations and compliance because you don’t have to modify the compliant code to run on a different platform and put that compliance at risk.
Some of the biggest names in banking and micro-services like FICO, Platform One, and Square have become huge fans of this technology, which assures IT managers that a problem is not due to the platform if it results in incompatibilities or catastrophic cloud outages.
One other exciting aspect of Tetrate is that it was designed as a post-pandemic company in that virtually everyone is remote. They have an office in Milpitas, California, for meetings, but employees are worldwide, including in the U.S., India, China, Indonesia, New Zealand, and Spain (at the time of this briefing, they were interviewing a new employee who lives in Israel). This position gives them far better access to the best talent because they don’t require employees to relocate.
The Emergence of an Inter-Cloud Hypervisor
Tetrate is at the front end of what should be a similar market opportunity to Hypervisors but, instead of providing portability across hardware platforms, Tetrate provides portability between cloud providers. Customers who use them in production praise this portability but also the reduction in management overhead and downtime. Security is also a central selling point as the solution directly mitigated the API exposure, which is forecast to be 90% of Gartner’s attack surface in a few years.
In the end, I can see why Dell and Intel invested in this company. Like VMware did in its day, they address a critical new problem in what is rapidly becoming a cloud computing world.