In this chapter, we describe different methods in which z/OS® users can use cloud object storage.

This chapter includes the following topics:

As explained in Chapter 1, “Cloud overview” on page 1, object stores are excellent target devices for cold or archive data. Content that often must be retained for longer periods and rarely accessed is a good candidate for object storage. Although thousands (if not millions) of devices in the world support the use of object stores, we focus on z/OS use cases.

Today, a few options are available for enabling the use of object stores under z/OS. Although some overlap exists between these options, each option includes some unique features.

The goal of this chapter is to highlight the differences between these offerings. We do not suggest which option is best for you. Because different options are always being added, this chapter is current as of this writing.

IBM Cloud Tape Connector for z/OS allows you to easily copy and move mainframe data to private, hybrid, or public Cloud storage. This feature offers improved security, flexibility, and economies of scale for archive or backup and recovery. IBM Cloud Tape Connector for z/OS processes and moves your data without the need for more hardware gateway devices. It uses zIIP processors to minimize CPU cost of data movement.

This offering provides IBM Z to cloud connectivity and is independent of the DASD or Tape technology that is used by the organization.

By emulating virtual tape devices, applications or utilities that can write to tape can use Tape Cloud Connector for z/OS. When content is copied or written to such an emulated tape, it is initially stored in DASD before being offloaded to an object store through a gateway device. Any access of the tape from that point forward requires that it be recalled back into DASD as a virtual tape device.

IBM Advanced Archive for DFSMShsm creates an archive tier for backing up inactive mainframe data that is managed by DFSMShsm. The added, lower-impact management tier uses less CPU resources to improve DFSMShsm efficiency and adds cloud storage to virtual and physical tape media options.

IBM Advanced Archive for DFSMShsm preserves DFSMS policies to properly back up data and manage retention. When applications or users recall data that is stored in a cloud, it is returned to DFSMShsm control and made available. IBM Advanced Archive for DFSMShsm supports IBM Cloud Object Storage, IBM SoftLayer®, and Amazon S3 cloud environments.

This offering provides direct IBM Z to cloud connectivity and is independent of the DASD or Tape technology used by the organization.

When a dataset is archived to the cloud, DFSMShsm surrenders awareness of the dataset after it is successfully stored in an object store. If this dataset is accessed, the advanced archive technology recalls the dataset back into DASD and again makes DFSMShsm aware of its presence.

For more information, see this website.

IBM intends to deliver a new cloud tier to OAM's existing storage hierarchy which will provide the ability to store and manage primary copies of OAM objects on cloud storage, via public or private cloud infrastructures supporting the Amazon S3 API, and the ability to recall an object stored in the cloud to the disk level of the storage hierarchy. OAM managed backup copies will continue to be supported as they are today to removable media, typically virtual or physical tape.

For more information see this website this website.

The IBM DS8000®, in combination with DFSMShsm, supports the ability to have datasets that are stored within the DS8000 be migrated to object store devices. Similar to how DFSMShsm supports ML2 (which often implies tape), DFSMShsm supports migration level cloud or MLC where the C stands for Cloud.

By modifying existing or new migration policies, DFSMShsm can request the DS8000 migrate a dataset from the DS8000 directly to a chosen object store through TCP/IP connectivity. No data movement through the host occurs and the dataset granular management eliminates the need for DFSMShsm recycle processing and other HSM inefficiencies, which saves CPU cycles on the host.

The DS8000 supports targeting different object store technologies and having the dataset objects migrate to the TS7700. The DS8000 essentially becomes a user of the TS7700 Grid and offloads datasets directly to the TS7700 through TCP/IP, which eliminates the need to migrate the dataset through the FICON® SAN.

For more information, see IBM DS8880 and z/OS DFSMS: Transparent Cloud Tiering, SG24-8381.

The IBM TS7700 supports the ability to have logical volumes tier to object stores, much as it has supported tiering to physical tape since the 1990s. Through partitions and policy management, logical volumes in their entirety can be premigrated to an object store and then eventually migrated or removed from TS7700 disk cache that is based on policy and LRU algorithms. This feature provides a tier of cold storage behind the TS7700, which improves the cost of ownership and redundancy of data within the TS7700. This IBM Redpaper publication is focused on this particular offering.

Although Transparent Cloud Tiering (TCT) is not a stand-alone product, it is important to understand what it is because it is mentioned frequently within the IBM portfolio.

TCT is an internal IBM software offering that enables numerous IBM products to use object storage. Instead of each product creating its own interface to many different object store vendors, TCT provides a common interface for IBM products to use across different object stores. This feature enables solutions, such as the DS8000 and TS7700, to have a much more consistent and simplified means of using object stores from different vendors that often can have variances in support and protocols.