The Metro/Global Mirror function enables a three-site, high availability disaster recovery solution. It combines the capabilities of both Metro Mirror and Global Mirror functions for greater protection against planned and unplanned outages. Metro/Global Mirror is supported across System z® and open-systems environments. Metro/Global Mirror is supported on the DS8000®. It is not supported on the DS6000™ or the ESS800.
Metro/Global Mirror is a three-site solution, which uses synchronous replication to mirror data between a local site and an intermediate site, and asynchronous replication to mirror data from an intermediate site to a remote site. In this configuration, a Metro Mirror pair is established between two nearby sites (local and intermediate) to protect from local site disasters. The Global Mirror volumes can be located thousands of miles away and continue to be updated if the original local site has suffered a disaster and I/O has to be failed over to the intermediate site. In the case of a local-site-only disaster, Metro/Global Mirror can provide a zero-data-loss recovery at the remote site as well as at the intermediate site.
Figure 1 shows the three sites that are used in a Metro/Global Mirror configuration. A Metro/Global Mirror environment is configured using a minimum of three storage units, one each at the local, intermediate, and remote sites. Data from the A volumes at the local site is synchronously replicated to B volumes at the intermediate site using Metro Mirror. Data from the B volumes at the intermediate site is asynchronously replicated to the C volumes at the remote site using Global Copy. FlashCopy relationships are created with the C volumes at the remote site as the FlashCopy source and the D volumes at the remote site as the FlashCopy target volumes, maintained as the consistent disaster recovery volumes using Global Mirror.
For Global Mirror, one storage unit at the intermediate site is designated as the master storage unit. The master storage unit sends commands over Fibre Channel Protocol (FCP) links and coordinates the consistency group formation process. These links are required for the Global Mirror master storage unit to coordinate the consistency group formation process with the storage units and to communicate the FlashCopy commands to the remote site. All status is relayed back to the master.
In a Metro/Global Mirror environment, if a PPRC pair between the local and intermediate sites is either duplex pending or becomes suspended, the Global Mirror session will fail consistency group formation. This is to ensure that the consistency group at the remote site is, in fact, truly consistent with the local site. There may be cases where the Metro Mirror secondary, at the intermediate site, is not aware that its status as a secondary is suspended, and consistency group formation would continue. The TSO RQUERY ACTION(DVCSTAT) or the API STAT4C query can provide information pertaining to the intermediate volume’s status in the Global Mirror session and both of the cascaded PPRC volume’s states (as a PPRC primary and as a PPRC secondary).