In the ever-evolving landscape of cloud computing, adaptability and flexibility are key. The features of Oracle Cloud VMware Solution account for these factors, and we’re thrilled to introduce a feature to revolutionize your VMware environment in the cloud: Multicluster software-defined data centers (SDDCs). This innovation provides unmatched versatility, cost savings, and workload separation capabilities, making your cloud journey even more efficient and cost-effective.
Traditionally, Oracle Cloud VMware Solution environments were limited to a single cluster within a single SDDC, which could be a bottleneck when it came to resource allocation and management. With Oracle Cloud VMware Solution being a customer-controlled environment, our customers were creating clusters manually with all the flexibility that the solution provides. To make this process easier for customers, we’re allowing customers to add up to six clusters within a single SDDC from the Oracle Cloud Console. This enhancement opens a world of possibilities for businesses of all sizes.
With the introduction of multicluster SDDCs, you can now create multiple clusters within a single SDDC. This feature offers many advantages, including the following examples:
Simplified management: With multicluster SDDCs, you get the best of both worlds. You maintain a single management pane comprising a single vCenter, NSX Manager, and HCX Manager to oversee all clusters within your SDDC, including simplified operations and centralized control, significantly reducing administrative overhead.
Workload separation: Each cluster within the multicluster SDDC can be dedicated to specific use cases or workloads. For instance, you can allocate a cluster exclusively for hosting databases, maximizing licensing benefits by licensing only the required cores. This separation ensures optimized performance and cost savings for diverse workload needs.
Cost efficiency: The ability to separate workloads not only optimizes performance but also maximizes cost efficiency. By licensing only the necessary cores for each cluster, you can significantly reduce licensing expenses, translating into substantial cost benefits for your organization.
Cluster variety: Multicluster SDDCs allow you to mix and match various Compute shapes within the same SDDC to suit your specific needs. For example, you can configure your unified management cluster with a dense shape, utilizing a vSAN datastore for virtual machines (VMs), while another cluster can be a standard shape, using OCI Block storage as a virtual machine file system (VMFS) datastore to expand your VMware environment. This flexibility ensures you can adapt your infrastructure to meet the demands of different workloads.
Scalability: Multicluster SDDC allows you to scale your VMware environment on demand. Build your ideal VMware cloud at scale, adding clusters from the get-go or seamlessly integrating them later. This feature gives you the flexibility to scale your environment up or down and out or in as needed.
By introducing the concept of multicluster SDDCs and building upon the numerous enhancements we’ve implemented in recent years, we proudly assert Oracle Cloud VMware Solution as the most scalable, versatile, and cost-effective VMware solution available in the cloud. Multicluster SDDC has the following technical specifications:
SDDC limits: A single SDDC can have up to six clusters. The unified management cluster must be the first cluster to be created in the SDDC.
Cluster types: An SDDC has the following types of clusters:
Unified management cluster: The unified management cluster is the primary cluster and is mandatory. This cluster hosts the management components for the SDDC, such as vCenter Server, NSX Manager, and HCX Manager, and allows you to host all your workload VMs, such as applications and databases. The minimum size for a unified management cluster is three hosts.
Workload clusters: Workload clusters are optional add-ons that you can create during or after SDDC creation to scale your SDDC out or in as needed. The minimum size for a workload cluster depends on the Compute shape you choose.
Dense: Three hosts
Standard: Two hosts
GPU: One host
Cluster limits: The maximum number of hosts in a cluster depends on the shape of the hosts in the cluster.
For dense and GPU shapes, the maximum number of hosts is 64.
For standard shapes, the maximum number of hosts is 32.
If a cluster contains a mix of shapes, the maximum number of hosts in the cluster is the lowest maximum allowed among all the chosen shapes. For example, a cluster with dense and standard hosts has a maximum of 32 hosts because standard shapes have the lower maximum allowed among the two shapes. To learn more about mixed-shape clusters, refer to Announcing the availability of Mixed-Shape Clusters for Oracle Cloud VMware Solution.
SDDC sizing: The size of each cluster depends on the type of workloads that you’re hosting. For example, if you’re hosting database workloads, you must choose a cluster with sufficient compute and storage resources. The Oracle Cloud VMware Solution Service has the following features:
Compute options: Intel, AMD, and NVIDIA processors with dense, standard, and GPU Compute shapes. Dense and GPU shapes have onboard NVMe drives, while standard shapes use OCI Block Storage as primary storage.
Network flexibility: Oracle Cloud VMware Solution seamlessly integrates into the same virtual cloud network (VCN) as your existing resources with minimal latency, top-notch security, and simplified management. With multicluster SDDC, we create all the underlying networking for you to either block or allow vMotion of workloads between the clusters.
Storage: Oracle Cloud VMware Solution offers a diverse set of storage options, including the following examples:
Dense shapes with onboard NVMe storage, which you can use as a vSAN datastore. Storage scaling is tied to host scaling when using vSAN.
OCI Block Storage, which can attach as primary storage for standard shapes or as secondary storage for dense shapes. This configuration allows for independent scaling of storage, separate from CPU and memory resources.
OCI File Storage service and OCI Object Storage to meet various file and object storage requirements
You can use multicluster SDDC in various use cases, including the following examples:
Separating databases: Databases often have different performance and security requirements than other workloads. By creating a dedicated cluster for your database workloads, you can ensure that your databases are getting the resources they need and that they’re properly isolated from other workloads.
Separating development, testing, and production workloads: You can use multicluster SDDC to separate development, testing, and production workloads. This separation can help to improve the security and stability of production workloads.
Meeting compliance requirements: Some compliance regulations require that certain types of workloads be isolated from other workloads. Multicluster SDDC can help you meet these compliance requirements.
Optimizing licensing costs: Multicluster SDDC can also help you to optimize your licensing costs. For example, you can create a dedicated cluster for your database workloads and license only the cores that you need.
Scaling out your VMware environment: Previously, Oracle Cloud VMware Solution customers could scale their SDDCs to a maximum of 32 hosts with standard shapes or 64 hosts with dense shapes. With the introduction of multiple clusters, a single SDDC can now scale anywhere from three hosts to up to 384 hosts!
SDDC with multi-vendor CPUs: Multiple-cluster SDDC allows customers to deploy different CPU-based clusters within the same SDDC. This reduces management overhead and simplifies day-2 operations. Previously, customers had to create a separate SDDC for each CPU-based compute shape. This could be time-consuming and complex, and it made it difficult to manage multiple clusters. With multiple-cluster SDDCs, customers can easily deploy and manage different CPU-based clusters within the same SDDC. This can save time and resources, and it can make it easier to scale and adapt your infrastructure.
Specialized workload clusters: Beyond Compute shapes, you can create specialized clusters, such as GPU-based clusters, to cater to high-compute requirement workloads. So, you can harness the power of GPUs for tasks like machine learning (ML), rendering, or scientific computing without impacting other clusters.
The following example shows how a customer could use multiple clusters to scale their VMware environment:
Each cluster is managed using a single vCenter Server instance, giving the customer a single point of control for all their VMware workloads. This example is just one of many of how multicluster SDDC can help create a customized VMware environment in Oracle Cloud Infrastructure.
Oracle Cloud VMware Solution’s multicluster SDDC feature empowers you to take your VMware environment in the cloud to new heights. With simplified management, workload separation, cost efficiency, cluster variety, and scalability, this feature offers a comprehensive solution tailored to your organization’s specific needs.
Whether you want to optimize your licensing costs, segregate workloads, or harness specialized clusters, multicluster SDDC is the answer. This new product enhancement allows you to optimize your VMware environment in the cloud with Oracle Cloud VMware Solution.
Praveen Kumar Pedda Vakkalam is a Principal Solutions Architect whose work is currently focused on Oracle Cloud Infrastructure and Oracle Cloud VMware Solution (OCVS). Prior to joining Oracle, Praveen was a Solutions Architect working primarily on various Storage products. His experience over the last 15 years was around architecting, planning, and implementation of large-scale data center migrations and hybrid cloud solutions.