Workspot is the first multi-tenant, cloud-native solution to deploy both VDI and Windows applications across multiple sites - including on-premises and public clouds like Microsoft Azure. Our obsession with customer success drives us to continually innovate and transform VDI and application publishing. With that in mind, I would like to talk about 3 key features today and the value they deliver to customers: (1) Cloud-managed RDSH Pools (2) Automatic Storage Management on Azure (3) Dynamic Feature Rollout.
Feature #1: Cloud-managed RDSH Pools
We are excited to announce the industry’s first cloud-managed, published RDSH application farms. Traditional remote-app architectures were designed in the client-server era where the “controller” and the RD session hosts were all running in the same LAN. As customers are migrating to multi-site architectures and a large number of customers are looking at easy cross-site failover or DR solutions, a new design for managing large, multi-site and multi-geo application farms is highly desirable.
We are extending our VDI architecture to manage RDS hosts directly from the Workspot Cloud. RDS hosts run the Microsoft remote desktop session host service. They are managed directly from the Workspot cloud using an agent. There is no “connector” in the middle. Each RD host runs an agent that talks directly to the Workspot Cloud over port 443. The architecture is location agnostic. RDS Hosts can run on any hypervisor and any location. This provides unprecedented flexibility to Workspot customers.
The legacy VDI 1.0 products use old client-server architectures where many of the components are statically setup for each set of VDAs. From an availability standpoint, the nested hub-spoke architecture of proxying traffic via a connector at every location is just bad design. It's a hack introduced by legacy vendors to retrofit on-prem products for cloud deployments. It's also a scalability bottleneck if every connection or every virtual machine-to-cloud API call is proxied through a windows VM in the datacenter. Additional work is required for sizing and HA and the promise by legacy vendors of simplicity is sabotaged again.
Instead, Workspot has leapfrogged legacy products by simplifying and optimizing the agent-to-cloud and cloud-to-agent bi-directional protocol over SSL/TLS. Users can be load balanced to the nearest datacenter in addition to addressing failover and DR scenarios.
Template and network fileshare replication best practices are already part of the Workspot Disaster Recovery solution when customers want to manage multiple sites for VDI DR today.
Setting up app-publishing or RD pools is simple:
- Set up the RDS host (Win2012 or Win2016)
- Install applications to be published, using any application installation tool
- Install the Workspot Agent. Convert VM into template or mark it as a master image for the RD Pool
- Create the RD pool on any hypervisor or Azure. Wait for all the hosts to go into "available" state
- Publish applications and assign apps to users or groups
When the process is complete, users get instant notification about new published applications.
One unique aspect of the Workspot architecture is that all devices are always connected to the Workspot cloud for policy updates over a latent control channel. We use a modern, full-duplex, web-sockets based communication mechanism to push instant updates to all the users globally. There is no need for users to re-login to see the new apps published in Workspot (unlike legacy products!).
Customers want simple and automatic failover as they think about hybrid deployments across multiple sites. The elegance of cloud products is directly proportional to the simplicity in the overall consumption experience and the fungibility in expanding or addressing more use-cases quickly. Having separate stacks for on-prem and cloud with no interoperability and migration path is a classic legacy vendor playbook to confuse the market.
Feature #2: Automatic Storage Management on Azure using Managed Disks
Data center admins need to provide a lot of information about infrastructure when creating VDI pools on hypervisors: (a) Which datastore or storage location should be used to create VM virtual disks? (b) Which virtual network (VLAN) should be used? (c) VM Customization information (like naming, AD membership, etc.)?
Then in order to ensure good performance and optimal capacity usage administrators need detailed plans: (a) Management of multiple resource locations (vSphere Clusters, ResourcePools, Datastores), (b) Type of disk for provisioning (full clones, thin clones, differencing disks, etc.), and (c) Storage limits handling (i.e., how many disks per volume?), etc.
Modern hyper-converged products hide this storage provisioning complexity. As a result, most IT teams expect the same simplified experience from public cloud infrastructure. And that is precisely why we're going to give them that experience - because it should be simple for customers.
Creating VDI pools on cloud infrastructure can be very different than creating them on hypervisors on-prem. Recently, Microsoft announced support for automatic storage management on Azure. Its Managed Disks feature introduces a layer of abstraction for disk management and hides all the limits previously associated with every storage account (such as 20,000 IOPS / account).
With Workspot, this Managed Disks feature means that you can now create up to 10,000 VM disks in a single subscription. Workspot customers only need to specify a template and type of storage (standard or premium) and clones are automatically placed on the right storage accounts.
Behind the scenes, the Workspot cloud plane provisions each VM with a managed disk and network interface using a simple pool-based naming scheme. The deletion or the update workflow (for non-persistent or RD pools) cleans up all resources, including virtual network interfaces and disks, automatically.
We have also added support for Azure tags. This simplifies the process of master-VM filtering for large deployments.
Best of all, this new feature is in production now. There is a single cost center; it's priced on a flat per-VM monthly rate. Workspot includes everything-- including the internal Azure compute, storage, networking and security elements, in a single SKU.
Feature #3: Dynamic Feature Rollout Delivers Innovation to Customers Fast
Another key feature is our ability to rapidly roll out new services to customers.
EUC cloud products are not only web SaaS. There are also a lot of native clients in the field, and they need regular upgrades and new features to leverage the cloud-first agility model. It is imperative that a cloud service regularly introduces new features and functionality in a controlled fashion to customers. That way, agility can be nicely balanced with high reliability.
The Workspot cloud control plane is designed to rapidly roll out new services to our customers as soon as they're ready. Multiple times per day, we deploy new services to our Dev cloud. And several times per quarter, we roll out well-tested new features to our Production cloud. Workspot modeled its approach to dynamic feature rollout for customers on the way modern cloud companies like Dropbox, Facebook and Google accomplish feature rollouts. We call our approach “Feature Egress".
Here's how it works:
- The main goal with this full stack instrumentation (cloud, clients, agents and connectors) is to be able to selectively turn on/off new features to tenants with fine-grained control.
- Most features span multiple components and are not part of a single code path. Having the ability to redirect users to different run-time behaviors and operating low-level infrastructure operations (like cloning on a hypervisor based on ops dashboard flags) enables us to scale and elegantly “gate features” to our production cloud.
- For example, a new feature that introduces functionality in the admin UI and Windows client can be exposed at a per-customer level, per-AD Group level, and also at a per-user level within a customer organization.
Continuous Integration/Continuous Development (CI/CD) tools are great for automating the process of shipping code to production. Most customers, however, expect controlled rollouts. They also expect some form of a heads-up for important new functionality in the EUC stack.
Workspot addresses these issues using “controlled rollouts," which means that all new major features are added as selective features. Here's how rollouts work:
- Features are divided into 3 segments: 1) low entropy, high value features, 2) high entropy, high value features, and 3) invasive and replace-the-old-functionality features. Based on the feature and the risk associated with it, the feature is either pushed out site-wide or a controlled rollout option is used. (Please note that we don’t ship low value features!)
- Customers are divided into numerous segments, including internal users, pilot customers, customers in early-stage, customers in specific verticals, etc. It’s similar to A/B testing in the traditional SaaS nomenclature. However, in the context of VDI the features can also influence the native-clients and other components in the overall picture.
- We monitor usage and use our unique Customer Success Engagement Model to collect early feedback from customers. A tight feedback loop is our core competency as a company and enables us to polish the functionality within a few months
Most importantly, we have added the functionality to track how the feature is doing in the field. This way, if any problems are encountered we can remediate or back-out the conditionals. All of these steps are completely automated (ops team dashboard driven). It took us several years to fine-tune the functionality across all our components.
A cloud-native, microservices-based pluggable architecture is the only elegant way to add this type of feature rollout because the cloud is a single service. This is a key assertion. We make it often because we want to really educate customers who might otherwise be bamboozled by cloud-washed products where a separate stack runs for every tenant.
List of current features in Selective Mode
As shown in the above screenshot, Workspot currently exposes "Azure Config” to select customers who are engaged with us to run Win10 VDI on Azure. The feature link allows us to enable the feature for the tenant and pass secondary controls to the customer IT team. This way, the customer can control how features are exposed to their end users also.
Shipping new features is great. However, if new functionality is not provided to customers in a timely fashion it defeats the cloud-first promise. The real test of a cloud-service is not the frequency of new releases every month. The real test is how quickly the new features are deployed to the customers in a controlled and non-disruptive fashion. Cloud DNA is not just about shipping software. It is also the machinery required to take the shipped feature to the customer end-to-end.
Workspot is transforming the industry
We are excited to announce that in addition to simplified management and infinite scalability, Workspot will also provide RDS-SALs for Workspot managed RD pools on Azure. Again, one cost center for all things related to VDI and app-publishing!
We care deeply about addressing the most important painpoints customers experience with legacy VDI solutions. The result is significant innovation in VDI that makes our customers more successsful, and the Workspot team is inspired to continue delivering on our commitment to customer success!