Migrating to Power Virtual Server on IBM Cloud

Backup and restore by using mksysb images

Migrating AIX workloads to Power Virtual Server by using the mksysb method offers a reliable and efficient way to ensure data consistency and system integrity. This method uses the well-established mksysb command, which creates comprehensive backups of the entire system, including the operating system, configurations, and data. The process minimizes downtime and disruption, allowing businesses to achieve a smooth migration from on-premises to cloud infrastructure. The mksysb method is highly customizable. Organizations can tailor backups to specific needs and optimize the use of cloud resources.

Security and performance are also key benefits of migrating to Power Virtual Server by using the mksysb method. Data encryption safeguards sensitive information both in transit and at rest, while isolated environments prevent unauthorized access. Power Virtual Server uses IBM® Power Systems' high-performance computing capabilities, helping ensure efficient operation of AIX workloads in the cloud. Enhanced networking features, including low-latency connections and high-speed data transfer, improve overall performance. Robust disaster recovery solutions and simplified management further ensure data protection, continuity, and ease of oversight, making this method a secure and effective choice for migrating AIX workloads to Power Virtual Server.

In this option, the existing AIX LPAR is backed up by using the built-in mksysb command. It might also require using savevg for LPARs with separate volume groups defined.

Consider this option for businesses that want an identical system that is migrated to IBM Cloud. It has the lowest cost and shortest time frame to migrate.

Migrating AIX workloads to Power Virtual Server by using mksysb and savevg provides a comprehensive approach that efficiently transfers all system and data elements to the cloud. This method uses AIX tools to create and transfer backups of the entire system, making it a reliable and well-structured migration strategy. By using these built-in commands, the migration process is streamlined, cost-effective, and time-efficient, with minimal downtime and disruption to business operations.

1. Create a mksysb backup on your on-premises AIX system to capture the entire root volume group (rootvg), including the operating system, configurations, and root-level data.
2. Transfer this backup to Power Virtual Server, where it is restored to set up the base system environment.
3. Use the savevg command to back up other volume groups beyond rootvg, capturing their respective data.
4. Transfer the backups to Power Virtual Server.
5. Restore by using the restvg command to check that all data and configurations are replicated accurately in the cloud environment.

This method is cost-effective because it uses existing AIX tools and eliminates the need for expensive third-party migration software. The mksysb and savevg commands are built into AIX, so you incur no licensing costs. Also, using cloud storage for transferring backups helps avoid the costs that are associated with physical media and dedicated migration hardware. The pay-as-you-go model of Power Virtual Server further reduces capital expenditures, optimizing overall migration costs.

In terms of time efficiency, this approach minimizes downtime through a streamlined backup and restore process. The mksysb command creates a comprehensive system backup in one go, reducing the complexity and time required for data transfer. Similarly, the savevg command ensures that additional volume group data is backed up efficiently. Transferring these backups through cloud storage expedites the migration process, ensuring a quick and seamless transition. This approach ensures business continuity with minimal disruption, making it a practical and efficient solution for migrating AIX workloads to Power Virtual Server.

Other benefits include the following points:

• Comprehensive backup and recovery: Ensures full system protection by backing up the entire root volume group (rootvg) and extra volume groups, capturing all necessary data and configurations.
• Cost-effectiveness: Uses built-in AIX tools with no additional software costs and takes advantage of cloud storage and the pay-as-you-go pricing model of Power Virtual Server to reduce overall costs.
• Time efficiency: Streamlines the migration process with comprehensive system backups and efficient data transfer, minimizing downtime and ensuring a quick transition.
• Flexibility and scalability: Offers tailored backups for specific system components and scalable resources to meet changing workload demands, optimizing the migration process.
• Enhanced data integrity and security: Maintains data integrity with accurate replication of the entire system and secure data transfer by using cloud storage.

Host and OS mirroring by using Geographic Logical Volume Manager (GLVM)

Migrating AIX workloads to Power Virtual Server by using Host and OS system mirroring with Geographic Logical Volume Manager (GLVM) helps ensure continuous data protection and high availability. GLVM replicates data at the logical volume level between on-premises systems and Power Virtual Server, minimizing downtime and maintaining business continuity. PowerHA SystemMirror enhances this setup by automating management and monitoring. This method uses Power Virtual Server's scalable and high-performance infrastructure, providing efficient data replication and resource optimization while offering robust security through data encryption and isolated environments.

This approach is cost-efficient, reduces infrastructure costs through a pay-as-you-go model and optimizes resource use. It provides robust disaster recovery solutions and ensures compliance with data protection regulations. The centralized control and simplified management that is offered by Power Virtual Server, combined with the automated processes of PowerHA SystemMirror, make the migration process streamlined and reliable. This ensures that businesses can efficiently move their AIX workloads to the cloud, maximizing the benefits of cloud infrastructure while minimizing risks and costs.

Database replication

Migrating databases to Power Virtual Server by using database replication offers the advantage of maintaining real-time synchronization between your on-premises database and the cloud environment. This method can offer minimal downtime and disruption to your business operations. By replicating data at the database level, any changes that are made to the source database are instantly reflected in the target environment, which helps maintain data consistency and integrity throughout the migration. This approach is beneficial for businesses that require high availability and continuous access to their databases, as it minimizes the risks that are associated with data migration and ensures a smooth handover.

The scalability and performance capabilities of Power Virtual Server allow optimization of resource utilization and adaptability to fluctuating workload demands. Power Virtual Server offers enhanced security features, including data encryption and isolated environments, which safeguard sensitive information and ensure compliance with data protection regulations. This method provides a robust disaster recovery solution, enabling quick recovery and continuity if of an on-premises system failure occurs. By using database replication for your migration to Power Virtual Server, you can enhance the resilience, flexibility, and overall efficiency of your IT infrastructure, while ensuring that critical business functions remain uninterrupted and secure.

Third-party replication software

Migrating AIX workloads from on-premises to Power Virtual Server can be efficiently managed by using various third-party applications. Options such as FalconStor StorSafe VTL, Double-Take® Move™ for AIX, and MIMIX are designed to facilitate seamless migration with minimal downtime and disruption. These tools ensure robust data integrity and continuous availability during the migration process, making them ideal for critical workload transitions. They offer features like high availability, disaster recovery, and flexible data seeding options, which streamline the migration and optimize resource utilization.

Using these third-party solutions, businesses can achieve a smooth migration to Power Virtual Server while benefiting from advanced backup and recovery capabilities. FalconStor StorSafe VTL, and Mimix for instance, optimizes the migration by emulating physical tape drives, while MIMIX is known for its near-zero downtime and realistic data testing. Each tool provides unique benefits that are tailored to different migration needs, ensuring data protection and high performance in the cloud environment. By using these applications, organizations can ensure a reliable and efficient migration, enhancing their IT infrastructure's resilience and flexibility.

AIX migration method limitations

Backup and restore by using mksysb images

Migrating to Power Virtual Server using a backup and restore approach with mksysb is straightforward and cost-effective, but it comes with several limitations. It requires system downtime, as it’s a point-in-time snapshot that doesn’t support incremental changes or live migration. The process only backs up the root volume group (rootvg), so extra data must be handled separately using tools like savevg. Post-migration, network, and device configurations often need to be manually adjusted to fit the cloud environment. It’s also less ideal for large systems due to potential performance and transfer bottlenecks, and it lacks the automation and real-time sync capabilities of advanced methods like GLVM or MIMIX. Despite these drawbacks, it’s a solid option for smaller, noncritical workloads where downtime is acceptable.

Host and Operating System Mirroring using GLVM

Migrating to Power Virtual Server using Host and Operating System Mirroring with GLVM offers high availability and real-time data replication but comes with several limitations. The setup is technically complex and requires expertise in AIX, LVM, and networking. It’s heavily dependent on high-speed, low-latency network connectivity, which can be costly and impact performance over long distances. GLVM replicates at the volume level, lacking application awareness and transaction-level consistency, making it less suitable for workloads that require fine-grained control. It’s also not ideal for small or one-time migrations due to its infrastructure and operational overhead. This method is best suited for environments that demand continuous availability and robust disaster recovery rather than basic workload transfers.

Database replication

Migrating AIX workloads to Power Virtual Server using database replication is a powerful approach for minimizing downtime and maintaining real-time data consistency. However, it has several important limitations to consider. First, database replication only transfers the data layer, it does not replicate the operating system, middleware, applications, or system configurations. The target environment in Power Virtual Server must be manually configured to match the source system, which can add complexity and risk if there are environment-specific dependencies. Also, the replication process often requires compatible database versions and schemas between the source and target, and sometimes, database upgrades might be necessary before replication can begin.

Setting up replication can also be technically complex, requiring deep knowledge of features like Db2 HADR, Oracle Data Guard, or third-party replication tools. These tools might have licensing costs and require specialized skills to configure and maintain. There’s also a potential performance impact on the source system, particularly in high-transaction environments, as real-time replication uses compute and I/O resources. During the final cutover, ensuring complete data synchronization and avoiding data drift can be challenging, especially if the source database remains active while switching to the target.

Overall, while database replication is ideal for scenarios where data availability is critical and application layers can be rebuilt or reinstalled separately, it is less suitable for full-stack system migrations or when an identical environment is required in Power Virtual Server. Proper planning, testing, and coordination are essential to avoid surprises during migration.

Third-party replication software

Migrating AIX workloads to Power Virtual Server using third-party tools (like FalconStor, MIMIX, Double-Take) can streamline the process and reduce downtime, but there are important limitations to consider. Licensing and subscription costs for these tools can be significant, especially for small or budget-constrained projects. Setup and configuration often require specialized expertise, and vendors might have differing support models or integration complexity. These tools might introduce compatibility issues with specific AIX versions, storage layouts, or custom configurations. Sometimes, third-party tools add infrastructure costs or require agents that impact performance. Vendor lock-in can be a concern if ongoing replication or HA/DR functions depend on the same migration tool. Finally, while these tools are great for minimizing downtime, they often need careful testing and validation to ensure data consistency and full system functions in the Power Virtual Server environment.

Migration process

1. Preparation
   • Assessment: Evaluated the current AIX environment, identifying all volume groups and critical data.
   • Set up Power Virtual Server: Provisioned the target environment in Power Virtual Server, ensuring compatibility with the existing AIX systems.
2. Create backups
   • mksysb backup: Generated a mksysb backup of the root volume group (rootvg), capturing the operating system, configurations, and root-level data.
   • savevg backup: Used savevg to back up extra volume groups, ensuring that all critical data was included in the backups.
3. Transfer backups
   • Cloud storage: Transferred the mksysb and savevg backup files to IBM Cloud Object Storage, using secure and reliable transfer methods.
4. Restore backups on Power Virtual Server
   • Provision VSI: Created a Virtual Server Instance (VSI) in Power Virtual Server to match the source AIX system's specifications.
   • Restore mksysb backup: Used the mksysb backup to restore the rootvg on the new VSI, setting up the base system environment.
   • Restore savevg backups: Applied the restvg command to restore the additional volume groups from the savevg backups, ensuring that all data and configurations were replicated accurately.
5. Testing and validation
   • System verification: Verified that the restored system on Power Virtual Server matched the on-premises environment, with all configurations and data intact.
   • Application testing: Conducted thorough testing of critical applications to ensure that they functioned correctly in the new environment.
6. Final cutover
   • Data synchronization: Ensured that the latest data changes were synchronized between the on-premises system and Power Virtual Server.
   • Production cutover: Switched production operations to the new environment on Power Virtual Server, with minimal disruption to business processes.
7. Post-migration monitoring
   • Continuous monitoring: Implemented monitoring tools to ensure that the new environment operated smoothly and addressed any issues promptly.

Results

Cost savings

• Infrastructure: By moving to Power Virtual Server, XYZ Corporation reduced the costs that are associated with maintaining and upgrading on-premises hardware.
• Resource usage: Using the pay-as-you-go model of Power Virtual Server optimized operational costs.

Operational efficiency

• Downtime: The streamlined migration process ensured minimal disruption to manufacturing operations.
• Performance: The new cloud environment provided better performance and scalability to meet growing data needs.

Data integrity and security

• Replication: The use of mksysb and savevg ensured accurate replication of the entire system and all data.
• Data transfer: Data was securely transferred and protected during the migration process.

By using mksysb and savevg, XYZ Corporation successfully migrated their AIX workloads to Power Virtual Server, achieving a seamless transition with significant cost savings and improved operational efficiency.

Backup and migration by using FalconStor VTL

For images over 2 TB, use FalconStor Storsafe VTL appliances to deduplicate and minimize the data transfer to the cloud. Images less than 2 TB can be transferred to IBM® IBM Cloud Object Storage using IBM® Backup, Recovery, and Media Services (BRMS) and IBM Cloud Storage Solution (ICC) IBM Cloud Direct Link is the preferred option, as it seamlessly connects on-premises resources to IBM Cloud, and provides a consistent, high-throughput connectivity without using the public internet.

Operating system-level backup and restore

Both the SAVSYS and GO SAVE commands can be used to backup the IBM® i operating system The GO SAVE command includes multiple options, three of which are discussed in this paper:

• Menu option 21: makes a backup of the entire system
• Menu option 22: makes a backup of all system data
• Menu option 23: makes a backup of all user data

Geographic mirroring can be utilized to migrate a host and operating system from an on-premises location to Power Virtual Server in IBM Cloud. Geographic mirroring occurs when all data in one environment is mirrored to a second location. There is both a synchronous and asynchronous version of geographic mirroring. Synchronous geographic mirroring over a distance might impact application performance due to communication latency.

Application-level replication

Several applications that are commonly deployed on Power have built-in replication capabilities:

• Db2 HADR
• Oracle Dataguard
• SAP HANA on-premises to cloud

Logical replication by using 3rd-party software

• Assure MIMIX (from Precisely): A high availability (HA) and disaster recovery (DR) solution for IBM® i, but it’s also widely used to support migrations — especially when businesses want minimal or near-zero downtime.
• Rocket iCluster: Logical replication software that runs on Power Systems running IBM® i. iCluster uses journaling to provide high availability and disaster recovery for business applications and enables almost continuous access through real-time monitoring, notifying, and self-correcting replication.
• Robot HA: A high availability and disaster recovery solution for IBM i that provides real-time data replication and failover capabilities, ensuring business continuity, and it plays a critical role in migration by enabling seamless cutovers to new systems with minimal downtime and data integrity preserved.

IBM® i migration method limitations

Falconstor VTL

Migrating IBM® i workloads to Power Virtual Server using FalconStor VTL is a reliable method that uses virtual tape backups to simulate traditional tape-based restore processes in the cloud. However, it comes with several limitations. It is a point-in-time backup and restore approach, meaning that it does not support real-time replication, and any changes that are made after the backup must be reapplied manually. This process typically requires planned downtime, making it less suitable for high-availability environments. The migration involves multiple manual steps, including backup creation, data transfer, and system restoration, and might require adjustments to system configurations in Power Virtual Server. Setting up FalconStor VTL also requires specialized expertise and proper network connectivity between the on-premises and cloud environments. For large workloads, data transfer times can be significant, even with deduplication. Also, it does not preserve active jobs, sessions, or in-flight data, which might impact business continuity during cutover. While FalconStor VTL is effective for full system migrations, it’s not ideal for environments that need incremental or near-zero downtime migrations.

Operating system-level backup and restore

Migrating IBM® i workloads to Power Virtual Server by using OS-level backup and restore, such as SAVSYS, GO SAVE menu options, or full system save (Option 21), is a traditional and cost-effective approach, but it also has several limitations. This method requires a full system shutdown or restricted state to run a consistent backup, resulting in planned downtime. It captures the system as a snapshot, meaning no real-time synchronization, and any changes after the backup must be reapplied manually. Also, it involves manual configuration steps during the restore process in Power Virtual Server, including setting up devices, network configurations, and IPL settings. Large environments might face long backup and restore times, especially if cloud storage or lower-bandwidth networks are used for transfer. The method also does not preserve active sessions or jobs, and doesn’t support automation or rollback features that are typically offered by replication tools. While suitable for smaller or noncritical systems, this approach might fall short for complex, high-availability environments that require minimal disruption and fast recovery.

Application-level replication (Db2, Oracle, SAP)

Migrating IBM® i workloads to Power Virtual Server by using application-level replication, such as Db2 HADR, Oracle Data Guard, or SAP HANA replication, offers targeted data replication with minimal downtime, but it has several limitations. This method replicates only the application data, not the operating system, user configurations, or the broader application environment. As a result, the target system in Power Virtual Server must be manually prepared, including OS setup, user provisioning, application installation, and configuration to match the source. It also requires that the replication-capable version of the application is installed and supported in both the source and target environments. Application-level replication typically involves complex setup, licensing, and infrastructure requirements, and might need expert knowledge to manage logs, failovers, and synchronization. Also, this method does not handle nonapplication files (for example, spool files, custom scripts, system objects), so full system functions might work without extra migration steps. While effective for business-critical databases or ERP systems, application-level replication is best used as part of a hybrid migration strategy, not as a complete solution on its own.

Logical replication with third-party software

Migrating IBM® i workloads to Power Virtual Server by using logical replication with third-party software, such as Assure MIMIX, Rocket iCluster, or RobotHA, is a powerful strategy for achieving near-zero downtime, but it has several limitations. First, these tools typically require separate licensing and ongoing costs, which can be significant, especially for small to mid-sized environments. Logical replication is journal-based, meaning that it relies on proper journal configuration and object selection; if not set up correctly, critical objects or data might be missed. These tools replicate libraries, files, configurations, and user profiles, but often exclude some system-level objects like licensed programs or temporary system data. Also, the setup is technically complex and might require specialized expertise in both the software and IBM® i internals. Successful failover and cutover require rigorous testing and monitoring, and while rollback is possible, it might not be as simple as with snapshot-based approaches. Finally, logical replication tools often introduce infrastructure costs and require stable, high-performance network connections between the source and target systems. Despite these challenges, this method is ideal for environments that require high availability and minimal disruption during migration.

Migration process

1. Preparation
   • Assessment: Conducted a thorough assessment of the existing IBM® i environment, identifying all critical applications, data, and dependencies.
   • Set up Power Virtual Server: Provisioned the target environment in Power Virtual Server, ensuring that it was configured to match the specifications of the on-premises IBM® i systems.
2. Create backups
   • Virtual Tape creation: Used FalconStor's StorSafe VTL to create virtual tape backups of the IBM® i systems. This included full system backups and incremental backups of critical data.
   • Deduplication: FalconStor's built-in deduplication technology minimized the amount of data transferred, reducing both time and resource requirements.
3. Transfer backups
   • Secure transfer: Transferred the virtual tape backups to IBM® IBM Cloud Object Storage using secure transfer methods to ensure data security and integrity. FalconStor's VTL ensured efficient and secure data transfer.
4. Restore backups on Power Virtual Server
   • Provision VSI: Created new Virtual Server Instances (VSIs) in Power Virtual Server to replicate the on-premises IBM® i environment.
   • Restoration: Used the virtual tape backups to restore the IBM® i systems on the new VSIs, reestablishing the operating system, applications, and data.
5. Testing and validation
   • System verification: Verified that the restored environment in Power Virtual Server matched the on-premises setup, with all configurations and data intact.
   • Application testing: Conducted extensive testing of critical healthcare applications to ensure functions in the new environment.
6. Final cutover
   • Data synchronization: Ensured the latest data changes were synchronized between the on-premises system and Power Virtual Server.
   • Production cutover: Transitioned production operations to the new Power Virtual Server environment, minimizing downtime and ensuring continuous availability.
7. Post-migration monitoring
   • Continuous monitoring: Implemented monitoring tools to ensure that the new environment operated smoothly and addressed any issues promptly.
   • Ongoing optimization: Continued to optimize performance and scalability to meet future business needs.

Results

Cost saving

• Infrastructure: By migrating to Power Virtual Server, LMN Healthcare significantly reduced the costs that are associated with maintaining and upgrading on-premises hardware.
• Resource usage: Used the pay-as-you-go model of Power Virtual Server to optimize operational costs and resource utilization.

Operational efficiency

• Downtime: The migration process was run with minimal downtime, ensuring continuous availability of healthcare services.
• Performance: The new Power Virtual Server environment provided improved performance and scalability to meet the growing demands of healthcare operations.

Data integrity and security

• Replication: FalconStor's VTL ensured accurate replication of the entire system and all data, maintaining data integrity throughout the migration process.
• Data transfer: Data was securely transferred and protected throughout the migration process.

By using FalconStor's StorSafe VTL, LMN Healthcare successfully migrated their IBM® i workloads to Power Virtual Server. The migration resulted in significant cost savings, enhanced performance, and improved operational efficiency.