Global Mirror
Global Mirror is asynchronous replication that operates between two sites that are over 300 km apart. The source volumes are on one storage system and the target volumes are on another storage system. Global Mirror replication maintains identical data in both the source and target volumes.
When a write operation is issued to a source volume, the changes are typically propagated to the target volume a few seconds after the data is written to the source volume. However, changes can occur on the source volume before the target volume verifies that it received the change.
Because consistent copies of data are formed on the secondary site at set intervals, data loss is determined by the amount of time since the last consistency group was formed. If the system fails, Global Mirror might lose some data that was transmitted when the failure occurred.
Global Mirror Single Direction
The Global Mirror Single Direction session type enables you to run Global Mirror replication from the primary site.
For IBM® DS8000® storage systems, each copy set in the single direction session consists of two host volumes and a journal volume. Figure 1 illustrates a Global Mirror Single Direction session for these storage systems.
TVT Instructions: This graphic was updated for Copy Services Manager 6.2.1 frg_i_global_mirror_v2.jpg. To capture this graphic from the Copy Services Manager GUI, click Sessions > Click Create Session > In the Create Session window, select DS8000, ESS 800 in the Choose Hardware Type list > In the Choose Session Type list, select Global Mirror Single Direction.
For FlashSystems/IBM Spectrum Virtualize and SAN Volume Controller storage systems, each copy set in the Global Mirror Single Direction session consists of two host volumes. Figure 2 illustrates a Global Mirror Single Direction session for these storage systems.
Global Mirror Failover/Failback
The Global Mirror Failover/Failback with Practice session type combines Global Mirror Failover/Failback and FlashCopy® capabilities to provide a point-in-time copy of the data on the secondary site. This session type provides volumes that you can use to practice for disaster recovery without losing your disaster recovery capability.
For IBM DS8000 storage systems, each copy set in the Global Mirror Failover/Failback session consists of two host volumes and a journal volume. Figure 3 illustrates a Global Mirror Failover/Failback session for these storage systems.
TVT Instructions: This is same graphic as for Global Mirror Single Direction at the beginning of this topic, frg_i_global_mirror_essds_v2.jpg.
For FlashSystems/IBM Spectrum Virtualize and SAN Volume Controller storage systems, each copy set in the Global Mirror Failover/Failback session consists of two host volumes. Figure 4 illustrates a Global Mirror Failover/Failback session for these storage systems.
TVT Instructions: This is same graphic as for Global Mirror Single Direction for SVC/Storwize storage systems at the beginning of this topic.
Global Mirror Failover/Failback with Change Volumes
Global Mirror Failover/Failback with Change Volumes replication provides the same capabilities as the Global Mirror Failover/Failback session type. However, Global Mirror Failover/Failback with Change Volumes sessions also provide the option of enabling or disabling the use of change volumes. Change volumes are available in FlashSystems/IBM Spectrum® Virtualize and SAN Volume Controller storage systems.
Change volumes are composed of a source change volume and a target change volume that contain a point-in-time image of the data from the source and target volumes, respectively.
A FlashCopy operation occurs between the source volume and the source change volume. The frequency of the FlashCopy operation is determined by the cycle period. For information about the cycle period and how it is set, see Global Mirror Failover/Failback with Change Volumes session properties. The data on the source change volume is then replicated to the target volume, and finally to the target change volume.
Because the data that is replicated between sites contains point-in-time changes rather than all changes, a lower bandwidth link is required between the sites. However, when you use change volumes, your data exposure can increase. Therefore, you might want to include or exclude change volumes depending on your network traffic or business requirements.
This session type is available only for FlashSystems/IBM Spectrum Virtualize and SAN Volume Controller storage systems. Figure 5 illustrates a Global Mirror Failover/Failback with Change Volumes session for these storage systems.
TVT Instructions: This graphic was updated for Copy Services Manager 6.2.1 frg_i_global_mirror_change_volumes_v2.jpg. To capture this graphic from the Copy Services Manager GUI, click Sessions > Click Create Session > In the Create Session window, select SVC or Storwize in the Choose Hardware Type list > In the Choose Session Type list, select Global Mirror Failover/Failback w/Change Volumes.
Global Mirror Failover/Failback with Practice
The Global Mirror Failover/Failback with Practice session type combines Global Mirror Failover/Failback and FlashCopy capabilities to provide a point-in-time copy of the data on the secondary site. This session type provides volumes that you can use to practice for disaster recovery without losing your disaster recovery capability.
For IBM DS8000 storage systems, each copy set in the Global Mirror Failover/Failback with Practice session consists of two host volumes, an intermediate volume, and a journal volume. Figure 6 illustrates a Global Mirror Failover/Failback with Practice session for these storage systems.
TVT Instructions: This graphic was updated for Copy Services Manager 6.2.1 frg_i_global_mirror_pract_essds_v2.jpg. To capture this graphic from the Copy Services Manager GUI, click Sessions > Click Create Session > In the Create Session window, select DS8000, ESS 800 in the Choose Hardware Type list > In the Choose Session Type list, select Global Mirror Failover/Failback w/Practice.
For FlashSystems/IBM Spectrum Virtualize and SAN Volume Controller storage systems, each copy set in the Global Mirror failover/failback with Practice session consists of two host volumes and an intermediate volume. Figure 7 illustrates a Global Mirror Failover/Failback with Practice session for these storage systems.
Global Mirror Either Direction
Global Mirror Either Direction enables you to run Global Mirror replication from either the primary or secondary site.
This session type is available only for IBM DS8000 storage systems.
The session consists of two host volumes and two journal volumes. Figure 8 illustrates a Global Mirror Either Direction session for these storage systems.
TVT Instructions: This graphic was updated for Copy Services Manager 6.2.1 frg_i_global_mirror_either_dir.jpg. To capture this graphic from the Copy Services Manager GUI, go to Sessions, and then choose Create Session. Then select DS8000, ESS 800 in the Choose Hardware Type list > In the Choose Session Type list, select Global Mirror Global Mirror Either Direction.
Global Mirror Either Direction with Two-Site Practice
Global Mirror Either Direction with Two-Site Practice enables you to run Global Mirror replication from either the primary or secondary site. This session type provides volumes on the primary and secondary site that you can use to practice for disaster recovery without losing your disaster recovery capability.
This session type is available only for IBM DS8000 storage systems.
The session consists of two host volumes, two intermediate volumes, and two journal volumes. Figure 9 illustrates a Global Mirror Either Direction with Two-Site Practice session for these storage systems.
TVT Instructions: This graphic was updated for Copy Services Manager6.2.1 frg_i_global_mirror_2site_pract_essds_v2.jpg. To capture this graphic from the TPC-R GUI, In the Create Session window, select DS8000, ESS 800 in the Choose Hardware Type list > In the Choose Session Type list, select Global Mirror Global Mirror Either Direction w/Two Site Practice.
Examples
Read the following scenarios for information about using Global Mirror for asynchronous, remote data replication between two sites.
- Global Mirror Single Direction
- Although Jane's FlashCopy and Metro Mirror copies were both planned, Jane realizes that sometimes a failure can occur, and she wants to ensure that the data is safe. Because Jane works in San Francisco, she wants her other site to be far away if a disaster occurs locally. Her other site is based in Houston. A minor earthquake occurs in San Francisco and power and communications both fails. Jane arranged for the data on customer accounts that recently opened or closed to be asynchronously copied in Houston, using Global Mirror. Jane risks losing the bytes of data that were being processed when the tremor disrupted the processing in San Francisco. However, she understands that it is minor inconvenience when weighed next to the value of backing up her data to an area that is not prone to earthquakes.
- Global Mirror with Practice
- Jane wants to run a Global Mirror with practice from San Francisco to Houston. She plans to verify her recovery procedure for the Houston site, but she cannot stop running the Global Mirror session while she takes time to practice a recovery. By using a Global Mirror with Practice session, Jane can practice her disaster recovery scenario in Houston while the Global Mirror session runs uninterrupted. When she practices running the applications at the Houston site, Jane is better prepared to recover the data if a disaster ever strikes the San Francisco site.
- Global Mirror Either Direction with Two-Site Practice
- Jane wants to run a Global Mirror with Practice from San Francisco
to Houston. She plans to verify her recovery procedure for the Houston
site, but she cannot stop running the Global Mirror session while
she takes time to practice a recovery. By using a Global Mirror Either
Direction with Two-Site Practice session, Jane can practice her disaster
recovery scenario in Houston while the Global Mirror session runs
uninterrupted. When she practices running the applications at the
Houston site, Jane is better prepared to recover the data if a disaster
ever strikes the San Francisco site.
Jane can use the Global Mirror Either Direction with Two-Site Practice session to run asynchronous consistent data replication from either the San Francisco site or the Houston site. She can practice the disaster-recovery tasks at the target site, regardless of the location of the current production site. Jane's business can run a consistent Global Mirror session from its Houston site back to San Francisco while it runs a production site at Houston.