Accessing remote IBM Storage Scale storage cluster in IBM Storage Fusion
You can configure a remote file system mount for IBM Storage Scale and access it. The container native data access is only for IBM Storage Fusion on On-premises, Bare Metal, VMware, and Amazon Web Services deployments.
Before you begin
- For remote filesystem of Amazon Web Services storage cluster, ensure that you deploy a Amazon Web Services storage cluster. For more information about the storage, see AWS storage cluster.
- For the remote mounted storage cluster, after its filesystem level is updated to 28.00 (5.1.4)
or later, it is recommended to enable
--auto-inode-limit
on its filesystem for IBM Storage Fusion. For more information about how to enable, see mmchfs command. - To access the container native data, you must first add a file system. To add a new remote file system connection, see Connecting to remote IBM Storage Scale file systems.
- To understand the number of file systems and clusters that can be remotely mounted, see IBM Spectrum Scale container native limitations.
About this task
You can run a container workload against an existing data set that is stored on an IBM Storage Scale file system. For example, an IBM Storage Scale file system contains lot of images and you want to run a container workload that specializes in image processing.
- The remote file system mount must be configured. For more information, see Connecting to remote IBM Storage Scale file systems.
- When the remote file system mount is created, the IBM Storage Fusion automatically creates a statically provisioned volume that points to the file system. You must know how to identify this Persistent Volume.
- Optionally, modify the file system path that is specified in the Persistent Volume so that it points to a specific directory in the file system, rather than the root.
- Now a Persistent Volume Claim (PVC) can be made against the Persistent Volume. It can be
accomplished by matching the
accessMode
and storage request against the specified in the Persistent Volume. - A workload that uses the Persistent Volume Claim can view the contents of the file system.
Configure the UID/GID to match the one that is used to write the data set on the IBM Storage Scale file system. If you do not configure, then the container workload cannot access the files, because it does not have the permission.