VMware Cloud Director Extension for Data Solutions is a plug-in for VMware Cloud Director that adds up data and messaging services to the VMware Cloud Director portfolio. It offers cloud providers and their tenants a set of data and messaging services such as VMware SQL with MySQL, VMware SQL with PostgreSQL, and RabbitMQ. In addition to deploying and managing those services at scale, tenants can use data analytics and monitoring through Grafana and Prometheus to check the health of a service and take action if an issue occurs.
How does the Data Solutions Extension work?
The Data Solutions Extension works in conjunction with Container Service Extension 4.0 or later. It helps providers publish popular data and messaging services to their tenants, who in turn can use them for building new or maintaining existing applications.
The services are deployed in a Tanzu Kubernetes Grid Cluster that is managed by Container Service Extension 4.0 or later. The latter plays a vital role in service deployment. There is a service operator that is installed in a selected tenant Tanzu Kubernetes Cluster which manages the whole lifecycle of the service from the inception to the dissolution.
Here is a high-level diagram of the Data Solutions Extension architecture.
Installation and Configuration Overview
Data Solutions Extension is deployed as an ova file and its plug-in gets connected with the Cloud Director instance from where it will be accessed.
There are other configuration steps along the way which must be performed. One such is setting up a container registry after the plug-in is installed. This configuration step is performed by the provider administrator and requires either using Tanzu container registry or the provider’s own. This registry will be used for pulling the Docker images of the supported services and the data solutions operator.
Another important step is for the data solutions operator to be installed in the Tanzu Kubernetes Grid Clusters where the data solution services will be deployed. The installation is made by a Tenant Data Solutions Operator user. The operator manages the creation, update, upgrade, and deletion of a service on behalf of the user.
Use Cases
The Data Solutions Extension can be used for various purposes by tenants like the creation of database and messaging services at scale. Once those services are available in a tenant org, every tenant user that has the needed permissions can create, upgrade, or delete a PostgreSQL, MySQL, or RabbitMQ service. Advanced settings are possible to be applied during the creation of a service, one such is enabling High Availability for VMware SQL with MySQL and VMware SQL with PostgreSQL.
In addition to the database and messaging service creation, tenants can also manage effortlessly their upgrade. Once a newer version is detected, the Data Solutions Extension interface will indicate that there is such available and prompt the user to take action. The tenant administrators or users can upgrade the selected service with just a click, this way preventing the respective service from vulnerabilities and keeping it in check.
Consumption Metering
After the deployment, comes the usage of the Data Solutions Extension. While it is free to be used by our partners, the services it deploys are considered billable and their usage is calculated per-core basis.
Note: Only containers in the Tanzu Kubernetes Grid Cluster that keep the deployed service will be considered billable and not the entire cluster or worker node.
VMware SQL with MySQL and VMware SQL with PostgreSQL
For VMware SQL services, all containers with data solutions operators and VMware SQL instances must be identified, and the used vCPUs by the services must be counted for reporting. The Configuration of vCPUs in Kubernetes is reported as milliCores. To calculate the number of vCPUs, divide the reported milliCores by 1000.
VMware RabbitMQ
Only configured Kubernetes vCPU Resource Limits for Containers with RabbitMQ Nodes must be identified for reporting. Configuration of Kubernetes vCPU Resource Limits may be reported as milliCores. To calculate the number of vCPUs, divide the reported milliCores by 1000.
The metering of the data solutions service is currently not automatic and requires some manual steps from the providers. For checking which containers run a VMware SQL or RabbitMQ service, you can use the Kubernetes CLI tool kubectl to identify them. For more information, check the VMware Product Usage Guide.
Another option for checking the service usage is by employing VMware Chargeback for checking the usage of services per tenant.
After the calculations are made, the provider must manually enter the numbers into the VMware Commerce Portal.
Consumption Reporting
For reporting the consumption of the Data Solutions services, the provider will first need to submit a one-off order in the Commerce Portal for all the data and messaging services they plan to use. The consumption of the deployed services will need to be filled in a separate row in the monthly billing order (MBO) in the Commerce Portal.
The data services which have HA enabled will have their replicas billable in case they are engaged in a hot disaster recovery site and billable only in case of disaster for cold DR sites. Please refer to VMware Product Guide for clarity on what hot and cold DR sites are.
Conclusion
Undoubtedly, the Data Solutions Extension is a long-coveted plug-in for VMware Cloud Director. It was created with the thought to deliver the Tanzu Data Solutions to Cloud Director. Now this is a fact for some services, though others still remain on the roadmap.
With Data Solutions Extension we aim at enriching the Cloud Director portfolio of services and enabling providers to deliver more services to their tenants which will boost their digital transformation.
For more information on the Data Solutions Extension requirements and deployment, check its product documentation. To find answers to questions regarding the deployment, upgrade, usage, and metering of the extension, read the following FAQ.
For any other questions, check our cloud provider Slack channel.