IBM Geographically Dispersed Parallel Sysplex
Today’s data centers must be resilient to handle planned and unplanned system outages without impacting the business of the enterprise.
Most solutions that provide continuous IT services availability are based on cluster server software, mirrored storage, and redundant networks. When a failure occurs, an integrated disaster recovery solution runs, in a pre-determined fashion, a set of tasks to restart the operating system, recover storage servers and networks, and restart applications on a different set of servers and storage in another location. This level of functionality is also defined as IT continuous availability.
IBM Geographically Dispersed Parallel Sysplex (IBM GDPS) is an automated solution that provides various levels of IT continuous availability through different implementation levels. GDPS, as a continuous availability and near-transparent disaster recovery solution, is based on Parallel Sysplex server functions and disk mirroring that uses Metro Mirror, Global Mirror. Some scenarios includes even software mirroring to enhance transparency of action to achieve IT goal set by business.
This chapter provides an overview of the various features that GDPS offers. For more information about GDPS, see GDPS Family: An Introduction to Concepts and Capabilities, SG24-6374, IBM System Storage Business Continuity: Part 1 Planning Guide, SG24-6547, and the following websites:
|
Important: GDPS is a set of continuous availability solutions for z Systems environments. The various IBM GDPS offerings are not products. They are delivered as IBM Global Service offerings.
|
This chapter includes the following topics:
27.1 GDPS solution offerings
GDPS is a family of offerings, for a single site or a multi-site application availability solution, with the capability to manage the remote copy configuration of storage systems, automate z Systems operational tasks, manage and automate planned reconfigurations, and do failure recovery from a single point of control.
GDPS is an integrated end-to-end solution that is composed of software automation, software, servers and storage, networking, and IBM Global Services to configure and deploy the solution, as shown in Figure 27-1. The GDPS solution has components in the areas that are denoted by dark shading.
Figure 27-1 Positioning of GDPS solution offerings
The benefits of GDPS are a highly reliable and highly automated IT infrastructure solution with a z Systems-based control point, which can provide robust application availability.
GDPS is enabled through key IBM technologies and architectures.
GDPS supports both the synchronous (Metro Mirror) and the asynchronous (z/OS Global Mirror or Global Mirror) forms of remote copy. GDPS V3.11 is the last release that supports Peer-to-Peer Virtual Tape Server (PtP VTS) for Remote Copying tape data. IBM Virtualization Engine TS7700 retains support for remote copying of tape data based on grid configuration. The GDPS solution is a non-proprietary solution, working with IBM and other vendors’ storage systems, if the vendor meets the specific functions of the Metro Mirror, Global Mirror, and z/OS Global Mirror architectures that are required to support GDPS functions.
GDPS automation manages and protects IT services by handling planned and unplanned exception conditions. Depending on the GDPS configuration that is selected, availability can be storage resiliency only, or can provide near-continuous application and data availability. Regardless of the specific configuration variation, GDPS provides a z Systems-based business continuity automation and integration solution to manage both planned and unplanned exception conditions.
To attain high-levels of continuous availability and near-transparent disaster recovery, the GDPS solution is based on geographical clusters and disk mirroring.
27.2 Components of a GDPS configuration
Figure 27-2 shows some of the many possible components that are supported in a GDPS configuration. GDPS is configured to meet specific client requirements. Not all of the components that are shown are necessary in all GDPS offerings or GDPS solutions.
Figure 27-2 GDPS integrates many components
The GDPS family of z Systems business continuity solutions consists of three major offering categories, and each category has several subofferings. Each GDPS solution is delivered through IBM Global Services, and is tailored to fit a specific set of client recovery requirements, budgetary requirements, physical distance and infrastructure, and other factors.
The following major categories of GDPS solutions are built on DS8000 Copy Services:
•GDPS/PPRC solutions that are based on Metro Mirror. Metro Mirror was formerly known as Peer-to-Peer Remote Copy (PPRC). GDPS PPRC comes in several versions:
– GDPS/PPRC, a Tier 7 solution
– GDPS/PPRC HyperSwap Manager, a Tier 6 solution
– GDPS/PPRC Multiplatform Resiliency for z Systems (also known as xDR)
– GDPS/PPRC Distributed Cluster management (DCM)
•GDPS/MTMM solutions that are based on IBM Multi-Target Metro Mirror. It is similar to GDPS/PPRC except that instead one, two synchronous copies are managed.
•GDPS/XRC solutions that are based on z/OS Global Mirror. z/OS Global Mirror was formerly known as extended remote copy (XRC).
•GDPS/GM (Global Mirror) solutions that are based on Global Mirror.
These GDPS offerings also can be combined into 3-site and 4-site GDPS solutions, providing Tier 7 recoverability in multi-site environments:
•GDPS/MzGM, based on z/OS Metro/Global Mirror
•GDPS/MGM, based on Metro/Global Mirror
|
Remote Copy Management Facility: Remote Copy Management Facility (RCMF), an offering that includes only the storage interface management functions of GDPS, is not available anymore. Therefore, there are no details and references to it in this chapter.
|
IBM GDPS active/active continuous availability is a solution for an environment that consists of two sites, which are separated by unlimited distances, running the same applications and having the same data with cross-site workload monitoring, data replication, and balancing. Based on asynchronous software replication, planned switches can be accomplished with no data loss (RPO 0), and when sufficient replication bandwidth is provided, the RPO can be as low as a few seconds if there is an unplanned workload switch.
Because IBM GDPS active/active continuous availability does not use storage system-based data replication, this chapter does not describe it in detail. For more information about this solution, see the documentation that is listed at the beginning of this chapter.
27.2.1 GDPS/PPRC overview
GDPS/PPRC is designed to manage and protect IT services by handling planned and unplanned exception conditions, and maintain data integrity across multiple volumes and storage systems. By managing both planned and unplanned exception conditions, GDPS/PPRC can maximize application availability and provide business continuity.
GDPS/PPRC provides the following attributes:
•A near continuous availability solution
•A near transparent disaster recovery solution
•A recovery time objective (RTO) of less than an hour
•A recovery point objective (RPO) of zero (optional)
•Protection against localized area disasters with a typical distance between sites that is limited to a 100-km fiber distance
The GDPS/PPRC solution offering combines z Systems Parallel Sysplex capability, IBM System Storage DS8000 Metro Mirror disk mirroring technology to provide a Business Continuity solution for IT infrastructures that have z Systems at the core. GDPS/PPRC offers efficient workload management, system resource management, and business continuity or disaster recovery for z/OS servers and Open System data, and provides data consistency across all platforms that use the Metro Mirror consistency group function.
The GDPS solution uses automation technology to provide end-to-end management of
z Systems servers, disk mirroring, tape mirroring, and workload shutdown and start. GDPS manages the infrastructure to minimize or eliminate the outage during a planned or unplanned site failure. Critical data is disk mirrored, and processing is automatically restarted at an alternative site if a primary planned site shutdown or site failure occurs.
z Systems servers, disk mirroring, tape mirroring, and workload shutdown and start. GDPS manages the infrastructure to minimize or eliminate the outage during a planned or unplanned site failure. Critical data is disk mirrored, and processing is automatically restarted at an alternative site if a primary planned site shutdown or site failure occurs.
GDPS/PPRC automation provides scalability to ensure data integrity at many volumes across hundreds or thousands of Metro Mirror pairs.
27.2.2 PPRC and HyperSwap
The HyperSwap function is designed to broaden the near continuous availability attributes of GDPS/PPRC by extending the Parallel Sysplex redundancy to storage systems. HyperSwap can transparently switch primary Metro Mirror storage systems with the secondary disk systems for a planned or unplanned reconfiguration. The HyperSwap function can reduce the time that is required to switch to the secondary set of disks while it keeps the z/OS systems active together with their applications.
HyperSwap can be implemented as part of the full GDPS/PPRC offering, or through the GDPS/PPRC HyperSwap Manager offering, which provides HyperSwap functions without the host system management capabilities of the full GDPS/PPRC offering.
Since GDPS/PPRC V3.2, the HyperSwap function uses the Metro Mirror failover/failback function.
For planned reconfigurations, failover/failback might reduce the overall elapsed time to switch the storage systems, thus reducing the time that applications might be unavailable to users.
For unplanned reconfigurations, failover/failback allows the secondary disks to be configured in the suspended state after the switch and record any updates that are made to the data. When the failure condition is repaired, resynchronizing back to the original primary disks requires only the copying of the changed data, thus eliminating the necessity of performing a full copy of the data. The window during which critical data is left without Metro Mirror protection after an unplanned reconfiguration is therefore minimized.
|
GDPS 3.11 is last release that will support disk subsystem which is not failover/failback capable.
|
27.2.3 GDPS Virtual Appliance overview
Not all customers run z Systems platforms with z/OS. GDPS Virtual Appliance provides the required levels of automation for planned and unplanned outages for those who run just Linux on z Systems or z/VM.
GDPS Virtual Appliance is a self-contained system that runs in a separate logical partition (LPAR) that has GDPS/PPRC code, which provides PPRC monitoring and management of replication. Additionally, the GDPS Virtual Appliance can manage z/VM systems that use production disks. For this purpose, the GDPS Virtual Appliance is working together with GDPS/PPRC xDR.
27.2.4 GDPS/PPRC Multiplatform Resiliency (xDR) for z Systems overview
With a multitiered architecture, there is a need to provide a coordinated near-continuous availability and disaster recovery solution for both z/OS and Linux on z Systems. The GDPS/PPRC function that provides this capability is called multiplatform resiliency for z Systems, and it can be implemented if the disks used by z/VM and Linux are count key data (CKD) disks. To provide these capabilities, GDPS/PPRC communicates and coordinates with System Automation for Multiplatforms (SA MP) that is running on Linux on z Systems.
Linux can run either of the following ways:
• As a guest under z/VM
• Native in a z Systems partition
GDPS/PPRC coordinates planned and unplanned HyperSwap for both z/OS and z/VM disks, providing continuous data availability spanning the multitiered application. It does not matter whether the first disk failure is detected for a z/VM disk or a z/OS disk; all are swapped together. For site failures, GDPS/PPRC provides a coordinated freeze for data consistency across z/VM and z/OS. Again, it does not matter whether the first freeze trigger is captured on a z/OS disk or a z/VM disk; all are frozen together.
27.2.5 GDPS/PPRC Distributed Cluster Management
GDPS Distributed Cluster Management (DCM) is a capability that allows the management and coordination of disaster recovery across clustered distributed servers and the z Systems workload (or workloads) for which GDPS is responsible.
DCM support is provided in GDPS/PPRC for both Symantec Veritas Cluster Server (VCS) clusters and IBM Tivoli System Automation Application Manager (SA AppMan). GDPS/PPRC can support both VCS and SA AppMan concurrently. DCM provides advisory and coordination functions between GDPS and one or more VCS or SA AppMan managed clusters.
DCM allows the management and coordination of planned and unplanned outages across non- z Systems distributed clusters in coordination with the z Systems workloads for which GDPS is responsible.
27.2.6 GDPS/XRC overview
z/OS Global Mirror, previously known as extended remote copy (XRC), is a combined hardware and z/OS software asynchronous Remote Copy solution for z Systems data. z/OS Global Mirror is designed to provide premium levels of scalability, reaching into the tens of thousands of z Systems volumes. Consistency of the data is maintained through the consistency group function within the system data mover (SDM). For more information about z/OS Global Mirror, see Chapter 22, “z/OS Global Mirror overview” on page 289.
GDPS/XRC includes automation to manage z/OS Global Mirror Remote Copy pairs and automates the process of recovering the production environment with limited manual intervention, including invocation of Capacity Backup (CBU), thus providing significant value in reducing the duration of the recovery window and requiring less operator interaction.
GDPS/XRC provides the following attributes:
•A disaster recovery solution
•An RTO of 1 - 2 hours
•An RPO of less than 2 minutes, and typically 3 - 5 seconds
•Protection against localized and regional disasters with unlimited distance between sites
•Minimal Remote Copy performance impact
27.2.7 GDPS/GM (Global Mirror) overview
GDPS/GM is similar to GDPS/XRC in that it supports virtually any distance; GDPS/PPRC is limited to 100 km. However, because of the differing characteristics of the underlying remote copy technology (Global Mirror), GDPS/GM extends the remote copy support to fixed-block architecture (FB) data.
GDPS/GM can be viewed as a mixture of GDPS/PPRC and GDPS/XRC. Just as GDPS/PPRC is a storage system-based remote copy technology, GDPS/GM is also disk-based, meaning that it supports the same mix of CKD and FB data that is supported by GDPS/PPRC. Also, being disk-based, there is no requirement for an SDM system to drive the Remote Copy process. And, like Metro Mirror, Global Mirror requires that the primary and auxiliary storage systems are from the same vendor; however, Global Mirror is only supported by IBM storage systems.
GDPS/GM resembles GDPS/XRC in that it supports unlimited distances between the application and recovery sites. Also, GDPS/GM does not provide any automation or management of the production systems; its focus is on managing the Global Mirror Remote Copy environment and automating and managing recovery of data and systems if there is a disaster. Also, like GDPS/XRC, GDPS/GM supports the ability to make Remote Copy copies of data from multiple sysplexes; each GDPS/PPRC license supports Remote Copy for a single sysplex.
In addition to its disaster recovery capabilities, GDPS/GM also provides an interface for monitoring and managing the Remote Copy configuration. This management includes the initialization and monitoring of the Global Mirror volume pairs that are based on policy, and performing routine operations on installed storage systems.
For more information about GDPS/GM, see GDPS Family: An Introduction to Concepts and Capabilities, SG24-6374.
27.2.8 GDPS 3-site solution overview
GPDS also supports 3-site configuration solutions by using a combination of the previous GDPS solutions. A 3-site solution can combine the advantages of metropolitan distance business continuity and regional or long-distance disaster recovery.
GDPS z/OS Metro/Global Mirror
GDPS/PPRC and GDPS/XRC are supported configurations for the 3-site configuration that is shown in Figure 27-3.
Figure 27-3 GDPS 3-site configuration with GDPS/PPRC and GDPS/XRC
The same primary volume for Metro Mirror and z/OS Global Mirror can be supported by two different GDPSs, a GDPS/PPRC for metropolitan distance and Business Continuity, and a GDPS/XRC for regional distance and disaster recovery. The two mirroring technologies and GDPS implementations work independently of each other, yet provide the synergy of a common management scheme and common skills.
You can obtain more information about GDPS 3-site configurations from your
IBM GDPS representative.
IBM GDPS representative.
|
z Systems data support: GDPS/XRC supports z Systems data for z/OS, Linux on z/VM, z/VM, and z/VSE. GDPS/XRC and GDPS/PPRC on the same volume is a z Systems solution only.
|
GDPS Metro/Global Mirror
You can use GDPS/PPRC with GDPS/Global Mirror (GDPS Metro/Global Mirror) to manage the configuration across all formats of data, as Global Mirror is not limited to z Systems formatted data.
GDPS Metro/Global Mirror combines the benefits of GDPS/PPRC that uses Metro Mirror with GDPS/Global Mirror that uses Global Mirror. A typical configuration has the secondary disk from a Metro Mirror Remote Copy configuration in turn becomes the primary disk for a Global Mirror Remote Copy pair. Data is replicated in a “cascading” fashion.
Combining the benefits of PPRC and Global Mirror, GDPS Metro/Global Mirror enables the following features:
•A HyperSwap capability for near-continuous availability for a disk control unit failure
•An option for no data loss
•Maintenance of a disaster recovery capability after a HyperSwap
•Data consistency to allow restart, not recovery, at either site 2 or site 3
•A long-distance disaster recovery site for protection against a regional disaster
•Minimal application impact
•GDPS automation to manage Remote Copy pairs, manage a Parallel Sysplex configuration, and perform planned and unplanned reconfigurations
In addition, GDPS Metro/Global Mirror can be used for both z Systems and open data, and provide consistency between them.
For more information about Metro/Global Mirror configuration, see Chapter 32, “Metro/Global Mirror configuration and setup” on page 361.
27.2.9 GDPS 4-site solution overview
There is two options covering 4-site solutions:
•MGM
•MzGM
The MGM 4-site configuration is managed by a combination of GDPS/PPRC and GDPS/GM in both the application and the recovery regions. It also supports the use of the incremental resynchronization function of Metro/Global Mirror that is extended to a new PPRC-XD relationship between the C-disk and the D-disk.
In MGM 4-site configurations, the application region is Region A and the primary disks are in Site1, that is, there is an ABCD configuration from Region A to Region B.
From a disk point of view, MGM 4-site is basically two MGM-IR 3-site configurations (ABC and CDA) plus a cascaded leg of PPRC-XD from the C-disks to the D-disks in the recovery region. It is a 4-site symmetrical mirroring configuration providing the same high availability and disaster recovery protection no matter which region is the application region.
The MGM 4-site configuration can manage incremental resync operations between the two sites in one of the regions or between the two regions. This capability is important in managing a 4-site replication solution when planning for the possibility of the loss of the intermediate site or even an entire region.
From a GDPS point of view, GDPS/PPRC manages PPRC between the two sites in the application region, and GDPS/GM manages the cross-region GM session. In MGM 4-site, another GDPS/PPRC is added into the recovery region. This GDPS/PPRC is used to manage the PPRC-XD leg from Site1 to Site2 in the recovery region, which can be converted to synchronous PPRC (for example, after a disaster or as the result of a region switch).
The MzGM 4-site configuration is a combination of GDPS/PPRC and GDPS/XRC. In first region, GDPS uses PPRC. Hyperswap technology is also implemented for high availability functions. PPRC is also used for the second region. Also in this situation Hyperswap is available. Between regions XRC is used. Incremental resychronization is available also between region 1 secondary disks and region 2, so in case of need to re-establish XRC session there is no need of full copy.
MzGM, similar to the MGM 4-site configuration, is somewhat two separate three sites MzGM configuration. The key advantages of using MzGM is nearly identical operation procedures when running in either region and most relevant copy when you want to perform scheduled disaster recovery tests.
27.2.10 IBM Global Services offerings for GDPS
The various GDPS offerings are not products. They are delivered as IBM Global Service offerings. GDPS is an end to end solution in which IBM Global Services tailors and installs the specific combination of components, integrated within the client’s environment. GDPS integration work, including education, is done by using IBM Global Services, in addition to the planning and installation of the code. This work ensures that the GDPS solution provides the intended value to all parts of the business and IT processes.
The following GDPS services and offerings are provided by IBM Global Services.
GDPS Technical Consulting Workshop
A Technical Consulting Workshop (TWC®) is a two-day workshop where IBM Global Services specialists work with your representatives to understand your business objectives, service requirements, technological directions, business applications, recovery processes, and cross-site and I/O requirements. High-level education on GDPS is provided along with the service and implementation process. Various remote and local data protection options are evaluated.
IBM Global Services specialists present a number of planned and unplanned GDPS reconfiguration scenarios with recommendations about how GDPS can assist you in achieving your objectives. At the conclusion of the workshop, the following items are developed:
•Acceptance criteria for both the test and production phases
•A high-level task list
•A services list
•A project summary
GDPS/PPRC HyperSwap Manager
IBM Implementation Services for GDPS/PPRC HyperSwap Manager helps simplify implementation by working with you to get GDPS/PPRC HyperSwap Manager and its prerequisites running with limited disruption to your business. Onsite planning, configuration, implementation, testing, and education help make your IBM Implementation Services for GDPS/PPRC HyperSwap Manager solution accessible in the most efficient manner. Because GDPS/PPRC HyperSwap Manager is a subset of the GDPS/PPRC offering it is good practice to review comparison between both of them.
GDPS/PPRC and GDPS/XRC
IBM Implementation Services for GDPS/PPRC or GDPS/XRC assists you with planning, configuration, automation code customization, testing, onsite implementation assistance, and training in the IBM GDPS solution. Either GDPS has capability to manage TS7700 grid configuration.
..................Content has been hidden....................
You can't read the all page of ebook, please click here login for view all page.