Why workloads rely on IBM Power

That’s why many organizations run these critical workloads on IBM Power. Built for environments where uptime, performance and resilience are non-negotiable, IBM Power has earned a longstanding reputation for reliability, availability and security. These qualities matter most when technology underpins core operations. 

It’s also why customers continue to rely on IBM Power for their most important workloads, from retail transactions and production management to warehouse operations and financial processing. These systems sit at the center of daily business activity, requiring the ability to scale securely while delivering consistent performance under heavy transactional demand. 

A true strategic advantage

Equally important is how these environments are managed. Service Express operates one of the largest privately owned IBM Power cloud in Europe, enabling organizations to consume resilience as a service. Customers gain high availability, disaster recovery and on-demand capacity without the burden of managing firmware, hardware life cycles or complex clustering. 

The result is greater strategic focus. Internal teams can spend less time maintaining infrastructure and more time optimizing the applications that differentiate the business — in turn, improving customer experience, accelerating innovation and ensuring compliance. 

For crown jewel applications in retail, finance, manufacturing and logistics, resilience isn’t a luxury; it’s the foundation the business relies on. Partnering with experts who understand this reality ensures these critical systems continue to perform without compromise. 

Discover why companies like Pall-Ex trust us with their crown jewels today. 

IBM i 7.6 arrives with strong security improvements

The most significant change in this release was the addition of multi-factor authentication (MFA). With this update, IBM i now supports MFA with common third-party solutions such as Duo, Microsoft and IBM’s own authentication tools. This brings IBM i in line with modern security practices. 

The big security improvements in IBM i 7.6 such as MFA, System ASP encryption and CFGHOSTSVR to enable only secure connections for the host servers are unique to 7.6.  Other enhancements to Navigator for i, DCM and SQL are also made available on IBM i 7.5 with a few even sneaking back to 7.4. 

Another area that saw steady progress was SQL development. The team responsible for SQL functions continued to add new capabilities throughout the year. These additions directly addressed user questions and requests, showing that IBM is listening to real-world needs. 

Power11 rolls out focused improvements

Shortly after introducing IBM i 7.6, IBM released Power11 hardware. For many customers, this was a welcome update, even if it didn’t deliver a dramatic performance jump over Power10. Instead, Power11 offers meaningful improvements in memory capacity, efficiency and system consolidation. 

One notable point was the lack of a new entry-level Power11 system equivalent to the Power10 P05 class. IBM chose not to refresh this smaller configuration. With Power10, the P05 and single socket P10 servers were limited to only 5 PCIe slots. This was a problem for multiple LPARs on the machine. In the P10 IBM i software tier, IBM released the 22B models with 2-sockets and thus had 10 available I/O slots.  When Power11 appeared, IBM effectively merged the 22A and 22B models and set the tier based on the selected processor. Up to 8 cores were P10 and above that P10 LPARs were limited to 4 processor cores.  

The S1012 was the last Power10 server announced. It will not be discontinued in July and may be the only model spared. Customers who need this level of capacity can still find value in existing Power10 systems, which are capable and widely available. However, IBM recently announced withdrawal from marketing for Power10, effective July 31, 2026. 

When comparing Power11 to Power10, the two share many similarities. Both use DDR5 memory, and the physical layout and expansion options remain largely the same. Where Power11 gains an edge is with higher memory speeds, more cores per socket and better energy efficiency. These improvements deliver more consistent performance across a range of workloads while maintaining IBM’s focus on uptime and reliability. 

AI and IBM’s Project Bob

AI was another topic people raised when talking about 2025. It’s not yet the main driver for IBM i shops, but it’s gaining attention. One of the biggest AI developments from IBM last year was Project Bob. 

Project Bob is an AI-first development partner designed to help developers write, test, modernize and secure software more efficiently. Unlike simple code suggestion tools, Bob works with an organization’s actual codebase and development standards and uses multiple AI models to deliver context-aware guidance across the software life cycle. Bob is currently in private tech preview and is already being used internally at IBM to boost developer productivity. 

For IBM i customers, this kind of AI tool suggests where the industry is headed, even if most workloads today don’t center around generative AI. Over time, tools like Bob may influence modernization efforts, development efficiency and integration of traditional platforms with modern software practices. 

What this means for 2026

The trends from 2025 show IBM i continuing on a steady path. Key improvements like MFA address real customer needs. Power11 offers the next step in system evolution with gains in capacity and efficiency. Interest in AI points to future innovation without it being the central focus just yet. 

For IBM i users, 2026 will likely be about making the most of these foundations. Whether that means planning upgrades, tightening security or exploring new capabilities, the platform is positioned for continued relevance and support. 

Listening before we lead

When we meet with you, the first conversation is never about a proposal or a platform. It’s about your world.

We invest time upfront to understand:

• Your role, responsibilities, remit and priorities
• Pressures and opportunities shaping their function
• Strategic business drivers, special projects and organizational change
• Operational challenges, constraints and desired future state
• Technology posture, risks, costs, resilience and performance gaps

This is a detailed, research-driven and face-to-face conversation built entirely around listening. We want to see what you see across the business, not just inside IT, so we can support your road map, not just solve isolated problems.

Turning insight into value

After the initial discovery, the real work begins. We help you pull everything together:

• Business context
• Technical realities
• Risks and constraints
• Pain points and opportunities
• Ideal outcomes and future state

This big-picture analysis enables us to identify where we can make meaningful contributions to closing gaps, reducing risk, improving resilience or creating a more cost-effective and future-proof environment.

Demonstrating we’ve truly heard you

This is where collaboration becomes visible.

We meet with you again and replay the narrative clearly, accurately and in your own words. This is one of the most powerful moments in our process. It demonstrates that we’ve listened, understood and internalised what matters to you.

It also builds:

• Credibility because we reflect the reality you live every day
• Trust because the conversation feels familiar and aligned
• Engagement because you see yourself in the story we tell

From here, we begin to open the door to the art of the possible, not by selling solutions, but by co-creating them.

Our clients feel the difference

“From the very first conversation, it was clear Service Express took the time to understand our priorities and constraints, engaging as a joint problem-solving team before offering any recommendations. Their approach validated our thinking, aligned with our business needs and helped us shape a clear path forward. It quickly felt like they were part of our internal team rather than an external supplier.” — Jezz Howell, Group Head of IT, Briggs Equipment UK Ltd. 

This perspective reflects the heart of our approach: a collaborative, consultative and insight-led partnership that feels like working with a trusted extension of your team. 

What does EOSL mean?

When IBM announces EOSL for a product, sometimes referred to as end of standard service (EOSS), it ends standard maintenance and development for that generation. This means that after January 31, IBM won’t provide service, updates or fixes and will cease developing new code updates, patches or fixes.

Your systems will continue to run after IBM ends support. However, losing access to critical updates and guaranteed service levels can introduce risks, such as slower response times or security vulnerabilities, over time. Planning now will help you avoid issues later.

Exploring options for your infrastructure

When your equipment reaches EOSL, your business has options. You can:

• Refresh your hardware: Upgrade to Power10 or Power11 to get full IBM support, performance gains and security updates. This option comes with higher upfront CapEx and requires downtime to complete the refresh.
• Move to the cloud: Rapidly scale workloads without upfront costs or additional resources needed for physical infrastructure. Take advantage of expert support, proven processes and scalable resources that adapt to your needs, helping your organization move faster and more confidently.
• Extend the life of your equipment: Leverage third-party maintenance (TPM) to prolong the useful life of your current assets. You’ll benefit from immediate continuity, predictable costs and little to no downtime, but won’t have the latest generation of hardware.

The right move for your infrastructure depends on weighing the advantages and disadvantages of each option above against your long-term business road map.

Considerations for your strategic road map

While EOSL can feel like a disruption to your operations, you can use it as an opportunity to reassess your long-term infrastructure strategy. Here’s what to consider when determining your next steps:

1. Extend your data center equipment life cycles

OEM refresh timelines often strain budgets and disrupt stable environments. Many organizations feel pressure to update hardware based on OEM milestones and support limitations, rather than performance data, which can lead to unnecessary spending and a decrease in equipment ROI.

Unlike cars or home appliances, most data center equipment doesn’t begin experiencing service issues immediately after the warranty expires. When your server, storage or network systems reach EOL/EOSL, failures may increase slightly. However, decades of data show that overall reliability declines gradually — not abruptly — for non-critical items, such as fans, power supplies and disks.

Extending the useful life of your equipment doesn’t mean keeping it past its use-by date indefinitely. Adjusting your current refresh cycle by one or two years is a more realistic expectation.

The end of sale stage alone isn’t enough to base your refresh decision. Instead, refresh triggers should focus on reliability, capacity, compatibility and performance factors, rather than simply the age of your equipment.

Take control of your refresh cadence by identifying what your business truly needs, rather than only relying on OEM milestones as your guide. Extending equipment life cycles is a data-driven strategy that lowers costs, reduces waste and improves asset ROI.

2. Explore alternatives to new hardware

When IT leaders hear “refresh,” they often think “buy new.” It’s important to note that there are alternatives to purchasing new, especially when budget pressures, supply chain delays and sustainability goals make that approach challenging.

Typically, OEMs recommend a refresh cycle of 3–5 years based on their business model. Conversely, most of our customers refresh every 5–6 years due to budget constraints, migration and deployment logistics. Refreshing isn’t a simple flip of the switch; it requires planning, timing and resources. Opting for alternative hardware options can extend your refreshes to 7–8 years.

With refurbished or previous-generation equipment, you can explore cost-effective alternatives and still meet performance and compliance requirements. Ultimately, you achieve the same results with lower CapEx and minimal operational disruption.

These real-world customer examples illustrate significant cost savings while maintaining comparable performance: 

The main message is to supplement your refresh plans with alternative hardware options, not to avoid buying new products or from the OEM entirely. Consider alternatives, such as N-1 generations and refurbished hardware, to meet capacity and performance needs at significantly lower CapEx. 

3. Adopt a hybrid maintenance approach

Choosing between OEM and TPM support isn’t an all-or-nothing decision. With a hybrid strategy, you don’t overpay for OEM maintenance, risk coverage gaps on EOL & EOSL equipment or refresh your hardware unnecessarily. 

By utilizing OEM support where it’s needed and TPM for hardware that is stable or past the OEM development cycle, the result is balanced performance, compliance and cost, leading to a more efficient long-term maintenance strategy. 

A hybrid maintenance model meets a variety of needs, but it requires a detailed assessment of your environment and regular oversight. Focus on finding a solution that aligns with your specific infrastructure support needs. 

Key takeaways for your optimization journey

By leveraging these three strategies, you can optimize your overall data center environment, improve service and achieve better CapEx and OpEx savings. 

Key takeaways for navigating the evolving IT landscape: 

• Real-world data support a measured approach to extending life cycles and considering cost-effective alternatives for refreshing. 
• End of sale is not the only trigger for a refresh; evaluate reliability, performance and capacity data first. 
• Refurbished or N-1 configurations can provide near-equivalent performance at a significantly lower cost. 
• Hybrid maintenance can optimize spend and uptime but requires clear asset management and governance to decide when to use TPM or OEM support. 

For a plan tailored to your environment, please contact our team. We’ll assess your asset portfolio and help develop a comprehensive data center modernization road map. 

Resilience isn’t redundancy: it’s design

Redundancy is having two of everything. It’s knowing that when one fails, your business doesn’t. That’s a subtle but critical difference. 

You can run multi-AZ, multi-region and even multicloud deployments, but if your failover plan exists only in a slide deck, you’re not resilient; you’re hopeful. And hope isn’t a strategy. 

True resilience means designing like the cloud will fail. It means asking uncomfortable questions like: 

• Can we run if our primary provider disappears?
• Can we spin up our critical workloads elsewhere, or are they tied to a specific ecosystem? 
• Do we actually test recovery, or do we merely claim to do so?

That’s where architecture matters. It’s not about chasing 100% uptime. It’s about engineering the ability to keep going when things go wrong. 

Resilience is messy, but it’s worth it

Here’s the hard part: resilience doesn’t come in a box. It’s not a checkbox in your cloud console. It’s layers of thought, design and discipline. It’s data replication, isolated recovery environments, hybrid strategies and yes, sometimes even good old-fashioned private infrastructure. 

Because resilience isn’t about avoiding the cloud or depending on anything to stay perfect forever. 

When AWS (or Azure, or Google or anyone else) stumbles, the businesses that stay online aren’t the ones with the most significant cloud bills; they’re the ones that are built for continuity. 

Architecting for resilience isn’t something you can toggle on. It’s about designing layers — physical, logical and operational — that can withstand a hit and continue to function. 

Start with data because it’s the hardest thing to replace. Replicate it intelligently, across regions or platforms, and understand the trade-offs between consistency, latency and cost. 

Then, examine the applications. Can they be restarted elsewhere without needing to rewrite everything? That’s where containerisation, orchestration and good dependency management earn their keep. 

Then comes infrastructure, and not just having a second cloud account but genuinely having a recovery path. That could mean a secondary site, a private cloud or a completely isolated recovery environment. The goal isn’t perfect mirroring; it’s survivability. If your architecture can’t run at reduced capacity while things recover, you’ve built convenience, not resilience. 

And finally, consider people and process. You can have the best failover tech in the world, but if no one knows when or how to use it, it’s theatre, not strategy. Practice it, automate it and break things on purpose occasionally. Real resilience is learned through controlled chaos. 

The takeaway

Every outage is a reminder that resilience isn’t optional. It’s a competitive advantage and the quiet confidence that when the internet catches fire, your business continues to serve customers as if nothing happened. 

It’s fine if your most critical applications live in the cloud. Just make sure they’re not living only there. Resilience isn’t about trusting the cloud less. It’s about trusting your architecture more. 

In this webinar, IBM i expert and IBM Champion Pete Massiello walks you through the essential steps to prepare for your next upgrade. From prerequisites and planning tips to post-installation requirements, this session will equip you with the knowledge and confidence to complete your upgrade correctly the first time. 

What you’ll learn:

• Understand what to check before starting your system upgrade 
• Learn which required PTFs to apply ahead of time to avoid issues
• Gain insight into post-installation steps that ensure a successful upgrade

Why isn’t sharing /QDLS as a NetServer file share a good idea?

Multithreading. NetServer is shipped as a multithreaded facility by default for optimal performance. Because /QDLS is an ancient directory technology that can’t support multithreading, any file share connections to /QDLS must be single-threaded.

Suppose an SMB client (e.g., a Windows PC) makes its initial SMB connection to the system via a share on /QDLS. The NetServer server-side connection will be a single-threaded job, and any additional accesses to other IFS directories, outside of /QDLS, will also funnel through that single-threaded connection.

Suppose an SMB client makes its initial SMB connection to the system via a share on a normal IFS share (“normal” is a share on a directory outside of /QDLS). The NetServer server-side connection will be a multithreaded job, and any requests to access /QDLS after that initial connection will fail. Therein lies the fundamental problem with /QDLS NetServer shares.

How do you avoid connectivity issues with shares on /QDLS?

The best practice approach is not to have a share on /QDLS in the first place. Suppose you have code running on your system that requires using the /QDLS folder. For example, you may have a population of old applications that execute CPYTOPCD commands to create ASCII files of database files in /QDLS. A good workaround is to use the /QDLS folder, but after you place files there, have your application move those files from /QDLS to a standard IFS directory outside of /QDLS. This allows the directory to be accessed via its own SMB share that a multithreaded connection on the server side will service.

If you must continue to use a share on /QDLS for whatever technical/application reasons and you’re experiencing the /QDLS connectivity issue described above, another approach — albeit not recommended for performance degradation reasons — is to configure your system to only use single-threaded server-side SMB connections with NetServer.

This approach ensures that client-side connections to a /QDLS share will always work because a single-threaded job on the system will always service the client’s connection attempts. Still, it’ll impact your system’s overall performance of IFS file shares. Single-threaded sessions that support file systems like /QDLS that aren’t thread safe are handled by prestart job QZLSFILE, and multi-threaded sessions that support file systems except /QDLS are handled by prestart job QZLSFILET. To force all SMB client connections to be single-threaded, you need to remove the QZLSFILET prestart job entry from subsystem QSERVER using these steps:

1. End NetServer: ENDTCPSVR SERVER(*NETSVR)
2. End the QZLSFILET prestart job: ENDPJ SBS(QSERVER) PGM(QZLSFILET) OPTION(*IMMED)
3. Remove the QZLSFILET prestart job entry: RMVPJE SBSD(QSERVER) PGM(QZLSFILET)
4. Restart NetServer: STRTCPSVR SERVER(*NETSVR)

NetServer will now use only the QZLSFILE prestart jobs (single-threaded) for all SMB client connections. Because the multi-threaded prestart job QZLSFILET entry was removed from the QSERVER subsystem, the QZLSFILET job cannot start when NetServer is restarted. Thus, all SMB connections will be forced into single-threaded connections, and /QDLS connection failures should no longer occur.

If you need to re-enable multithreading for NetServer SMB connections, then execute these steps:

1. Add the QZLSFILET prestart job entry back to subsystem QSERVER: ADDPJE SBSD(QSYS/QSERVER) PGM(QSYS/QZLSFILET) STRJOBS(*NO) INLJOBS(1) THRESHOLD(1) ADLJOBS(0) MAXJOBS(*NOMAX) JOBD(QSYS/QZLSPJ) MAXUSE(1) CLS(QSYS/QPWFSERVER)
2. Start the QZLSFILET prestart job: STRPJ SBS(QSERVER) PGM(QZLSFILET)

Many shops still use the /QDLS directory in their application/processing environments; however, sharing the /QDLS directory using NetServer has some caveats, and hopefully, you now have some good insight into what those may be.

1. Invest in building a skilled team

Technology can detect data center issues, but it’s up to knowledgeable employees to resolve them. Start by hiring people with the right skills and experience for your environment. Since human error can cause unplanned downtime, having an IT team that can keep pace with changing technologies is crucial, so be sure to invest in staff training and certification. 

2. Consider outsourcing to third-party providers

While in-house teams are essential to IT operations, they may lack the skills, expertise or capacity to complete every maintenance task. Outsourcing responsibilities to a TPM provider helps organizations maintain performance without overwhelming internal staff or skipping maintenance due to limited time or resources. Depending on your IT needs and service level agreements, you can leverage OEM maintenance or a combination of OEM and TPM support. 

3. Maintain a targeted spare parts inventory

Having direct access to critical spare parts helps decrease work order delays and can enable quicker repairs in certain cases. Implement a virtual inventory system that tracks parts across multiple locations to streamline your routine maintenance, resolve unexpected or urgent needs and minimize downtime. 

4. Monitor energy consumption and carbon impact

Meeting sustainability compliance and regulations contributes to long-term data center success and resilience. It’s increasingly important for our planet’s health to adopt environmentally conscious practices, including using renewable energy and reducing carbon impact. At the same time, monitoring energy use prevents outages and extends the useful life of equipment. Effective energy management is both a compliance necessity and a key strategy for your infrastructure. 

5. Keep a clean, stable environment

A tidy and controlled environment is important for maintaining your data center equipment. Regular cleaning, like dusting and sweeping, prevents hardware from overheating, general wear and tear and debris that can create fire hazards. Keep temperatures and humidity levels stable to lower stress on hardware and extend its lifespan. Monitoring tools and cooling solutions can help improve indoor conditions and reduce maintenance issues or outages. 

6. Implement an asset management system

Efficient asset management in a data center depends on meticulous record-keeping and organization. Tracking the status, maintenance history, and configurations of servers, storage devices, networking gear and other hardware allows quicker troubleshooting and proactive maintenance. 

7. Create redundancies to ensure uptime

By implementing multiple redundancies, businesses can maximize uptime during equipment maintenance or unexpected outages. With additional servers, power supplies, cooling units and network connections as backups, you can minimize disruptions when primary equipment fails and make routine maintenance safer and easier. 

8. Establish testing protocols

Testing is crucial for identifying when maintenance is needed and confirming that it was successful. Developing testing protocols for key systems provides clear benchmarks to guide maintenance decisions and verify outcomes. High-priority systems — like backup power, fire suppression and emergency response equipment — require regular testing. These steps help ensure system reliability when you need it most. 

9. Prepare for emergencies

Even with strong data center maintenance practices, unexpected outages and disasters are sometimes unavoidable. Minimize disruptions by creating a comprehensive disaster recovery (DR) plan that details steps to take during a crisis, such as activating backup power, notifying vendors and protecting essential infrastructure. Regular drills, equipment inspections and proactive risk assessments help teams prepare for and prevent minor issues from turning into major failures. 

10. Safeguard your physical and digital environments

Security is an increasing concern for many businesses, including physical access to data centers and cyber threats. Physical security measures, like access control systems, surveillance cameras and intruder alarms, help prevent unauthorized entry and safeguard the facility from bad actors. On the digital front, backup and recovery services and cybersecurity tools protect sensitive data from cyberattacks and malware. Proper data storage and management minimize risks and help prevent failures, ensuring your operations’ reliability and longevity. 

Our data center maintenance solutions

Regardless of your data center maintenance strategy, following best practices and staying on top of your environment will help keep your operations running safely and reliably while saving your IT team time, budget and resources. Explore our Data Center Maintenance services to learn more about how Service Express can help support your infrastructure.