Hyperswap and Metro Mirror features of IBM PowerHA Systemmirror

Brief introduction

Data center and service availability is one of the most important topics in the IT infrastructure and attracts more attention every day. Data replication between sites is a good way to minimize business disruption because the backup restore operation takes too long to meet business requirements, or the device may be corrupted and cannot be used for data recovery, depending on the severity of the disaster. Recovering from the cheapest (recovery takes a long time) to the most expensive (providing the shortest recovery time and the closest 0 data loss) restores the cost of recovery options.

PowerHA systemmirror 7.1.2 Enterprise Edition provides a disaster recovery and high-availability solution that helps automate node failures and application events and provides high availability. It helps automate recovery of storage failures for selected storage, control storage replication between sites (separate data centers), and support the recovery of entire site failures, ensuring that replicas are in a consistent state, enabling failover and enabling you to build a disaster recovery solution.

Hyperswap is a product in the PowerHA systemmirror 7.1.2 Enterprise Edition product portfolio. This is a feature that is provided for storage errors and is provided to achieve continuous availability. It is based on stored synchronous replication [peer-to-peer Remote Copy (PPRC) or Metro Mirror]. When directed (or if a disk fails), the IBM AIX host that accesses the primary disk subsystem can transparently switch to the backup copy of the data, so that the user of the disk, such as the middleware, is unaffected.

Background knowledge

Hyperswap is the first feature introduced in GDPS a few years ago and is suitable for the Metro Mirror PPRC (sync) environment, which facilitates the immediate switching of PPRC mirrored disk subsystems, thereby enhancing Parallel Sysplex resilience.

If there is a physical connection between the host and the secondary storage subsystem, then the HYPERSWAP technology allows the host to transparently switch the application's I/O operations to the secondary Metro Mirror volume. This allows users to provide continuous operations from a single site or from multiple locations covered. By implementing Hyperswap, you can perform disk failure and maintenance functions without any application service disruption.

The solution provides customers with a better disaster recovery solution and proves the tight integration of PowerHA with IBM storage.

Hyperswap technology enables PowerHA Systemmirror to support the following functions for customers:

Make the primary disk subsystem not a single point of failure, providing a higher level of continuous operational support within the scope of coverage.

Support for storage maintenance without any application downtime.

Support for migrating from old to new storage.

All of these use cases are divided into two types of HYPERSWAP activities:

Unplanned Hyperswap: When primary storage fails, the operating system of the managed application detects events and reacts and performs PPRC failover, so the application's I/O activity is transparently redirected to the level two storage subsystem to allow applications to continue without interruption. Note that in this case, the operating system's Small Computer system Interface (SCSI) disk driver detects an error and makes a decision across multiple hosts to completely switch to the secondary storage subsystem. During the Hyperswap exchange, further processing of the I/O activity is temporarily frozen. Note that the application does not encounter a failure during this time, but instead encounters a non-fatal delay.

Planned Hyperswap: In this case, the administrator initiates a hyperswap from the primary storage subsystem to the secondary storage subsystem. When an administrator requests the execution of planned hyperswap, it coordinates the hosts in the cluster and then freezes the I/O activity. The exchange is performed first, and then the I/O operation is allowed to continue. Planned Hyperswap helps to perform maintenance tasks on primary storage and helps migrate data from old storage to newly purchased storage subsystems.

Figure 1:powerha Systemmirror Hyperswap Configuration Example

See more highlights of this column: http://www.bianceng.cnhttp://www.bianceng.cn/Programming/extra/