Stratoscale Launches Chorus As On-Site AWS

In a cheeky lead-in to AWS re:Invent next week, hybrid-cloud infrastructure company Stratoscale has launched Chorus, a suite of AWS-compatible services for running cloud-native applications, only on your own site.
Chorus is a fully-managed service that sits on top of Stratoscale's Symphony IaaS services, which are also designed to be AWS compatible.

Stratoscale have assembled an impressive list of services for launch: Load balancers (compatible with AWS Elastic Load Balancing), object storage (AWS S3 compatible), file services (Elastic File Services compatible), and database services (Relational Database Service compatible), even a managed Kubernetes option supporting Kubernetes 1.8.1.

S3 has become the de-facto standard for object storage, and there are plenty of vendors that offer S3 compatible products (Cloudian, Scality, NooBaa, Hedvig) or an interface to S3 storage (every enterprise backup vendor, most of the enterprise storage vendors). AWS lagged for some time in providing its own native file services and there are virtual storage appliance version of most storage vendor products that do file storage (NetApp, HPE, Dell EMC), and load balancers have been available in software form for ages (F5, nginx, etc.). To have all of them, you need to assemble them all together to match the full suite provided by AWS.

What I find striking is that Stratoscale is attempting to do what OpenStack arguably failed to do: provide the full AWS-like cloud native experience without being AWS. OpenStack tried to be AWS only better, while Stratoscale is providing a kind of on-site AWS Greatest Hits. OpenStack is definitely popular within its own niche, but it failed to capture the imagination of most developers. I recall with some sadness the early debates of whether OpenStack should build AWS API compatible services or not. Look at the number of SWIFT-based storage services compared with those using S3 to see the outcome.

“We provide what customers have told us are the core functions of AWS,” said Stratoscale CEO Ariel Maislos. “We will lag AWS features, but we have a very fast roadmap, and we give customers visibility of what we're planning six to twelve months ahead.”

For most customers, this is likely to be sufficient. If you need cutting edge features, or a niche function, AWS is still there for you to use, but for a lot of enterprises, the core functions of a service-based architecture is what they want. There may be reasons the cloud-based AWS isn't quite right for them: network bandwidth, or data sovereignty issues, for example.

“With Stratoscale Chorus, you can have an AWS region in your own datacenter,” said Maislos.

I see Chorus as complementary to AWS. There may be some customers who are happy with “Just Enough” AWS, but I think most would use Chorus to supplement existing or planned AWS usage to fill in some gaps in AWS' ability to run on-site. Some long-running low-variable workloads are cheaper to run on your own infrastructure, while AWS is excellent for more variable or bursty workloads.

The challenge will be if Chorus is compatible enough. AWS is constantly changing things and adding new services, but if Stratoscale are right about their choice of what customers consider core services—and price them correctly—I can see the product doing well.

Of course, AWS might well develop its own on-site deployment option that would immediately threaten Stratoscale's entire reason for existing, but we've seen how challenging that is with Microsoft's adventures in AzureStack, or perhaps AWS still believes that its vision of public cloud is the One True Way and won't even bother to try to build one.

I'll certainly be watching the announcements next week to see if AWS steals Stratoscale's thunder.

This article first appeared in Forbes.com here.