Planning for IBM Storage Scale on public clouds

Learn about the planning considerations for using cloudkit on public clouds.

After going through the planning considerations, see the following sections for installation, configuration, cloudkit command options, and cloudkit issues:
Note: Currently, the cloudkit supports to deploy an IBM Storage Scale cluster on the public cloud of either Amazon Web Services (AWS) or Google Cloud Platform (GCP). For more information, see Supported features of cloudkit .

IBM Storage Scale on cloud: Support statement

IBM Storage Scale deployments and functions are supported on a variety of clouds, directly through IBM, and also through Business Partners who have crafted specific solutions.

Recommended: Use of the IBM Storage Scale cloudkit for deployments is highly recommended, because each option and configuration parameter has been vetted against best practices, thus resulting in a known and workable cloud configuration. If the cloudkit does not support your use cases, refer to the alternate methods for provisioning.

Alternatively: Manually provisioning cloud resources and deploying IBM Storage Scale without using the cloudkit is also supported on the following clouds.

  • Amazon Web Services (AWS), Azure, IBM Cloud, Google Cloud Platform (GCP)
    Restriction: IBM Storage Scale consists of many features, some of which are highly useful and relevant for cloud environments, while other features are not recommended in a cloud environment. A few examples:
    • SMB, NFS, Object protocols using Cluster Export Services (CES stack). The CES stack relies upon existence of an assigned IP pool (non-DHCP) that can freely rotate across nodes. The IBM Storage Scale CES stack controls assignment and rotation in cases of node failure. In a cloud environment, IPs can only be assigned and moved by using cloud-specific APIs; which means that the IBM Storage Scale CES stack is currently incompatible with cloud APIs without further development.
    • AFM home (using NFS protocol). See the previous CES stack explanation, because of the CES NFS protocol is used for AFM home.
    • Cloud block disk services are a feature of every cloud and provide block access to storage. IBM Storage Scale uses disk services to create NSDs for building the IBM Storage Scale file system. For compatibility to be ensured, these disk services must be fully vetted with IBM Storage Scale. An inclusion list of services specifically compatible with IBM Storage Scale is shown in the supported features matrix for each cloud.
    • Custom Kubernetes services are a feature of every cloud. IBM Storage Scale supports manually installed Vanilla Kubernetes with its CSI driver. Custom services such as EKS and GKS are not supported by the CSI driver.
    • OpenShift is available in every cloud as either a self-managed instantiation or a bundled service. Refer to the IBM Storage Scale Container Native Storage Access (CNSA) documentation for a list of supported clouds for OpenShift.
    • Virtualization technologies on the cloud: VMware/KVM. The IBM Storage Scale FAQ contains a compatibility matrix of features and functions supported in VMware/KVM. This same matrix applies to VMware/KVM on the cloud.
    • For other exceptions, as well as lists of supported features, see the supported features matrix of IBM Storage Scale for each cloud.

If you choose not to use the cloudkit, it is your responsibility to make sure that you adhere to all restrictions and best practices in the rest of the IBM Storage Scale documentation.

If your are using a Business Partner’s cloud deployment of IBM Storage Scale, contact the Business Partner for support statement details.