For most IT leaders today, efforts to eliminate downtime for their organization mostly focus on unplanned downtime – removing downtime that is unexpected and not scheduled. This is because the majority of organizations offer SLAs (Service Level Agreements) that promise an uptime of up to 99.99%. However, in the current world of native-Cloud environments, the issue of unplanned downtime is no longer the threat it previously had been to legacy applications living in on-prem environments.
The more imminent type of downtime, however, which most organizations don’t give as much thought to, is planned downtime. Planned downtime is essential to performing maintenance for such things as periodic updates, security patches, and more. In these instances, systems need to shut down in order to perform the necessary upkeep of their IT estate.
Strategies to reduce all outages for an enterprise should take into account both types of downtime, and the number of investments made in tools and resources to avoid unplanned downtime should be equally made in avoiding planned downtime. This is critical because limiting planned downtime requires completely different methodologies. It is important to keep in mind that both kinds of downtime can be equally devasting to an organization’s reliability, revenue, and reputation – but at the same time, each can also be diminished with the appropriate approach.
This blog will provide an overview regarding why planned downtimes can be just as devastating to a business as an unplanned outage, and will also include some helpful suggestions regarding how you can mitigate your enterprise IT’s planned downtime.
Planned Downtime Can Mean High Costs to Your Organization
Even though both planned and unplanned outages can be equally harmful, much more consideration has been dedicated to avoiding unplanned downtime than to avoiding planned downtime.
Here are two key misconceptions for why unplanned downtime is considered more threatening than planned:
1. The misconception that unplanned downtime is more costly than planned.
For years, there has been the belief that unplanned downtimes are much more costly than planned ones because planned downtimes can be better controlled and scheduled for periods when there is less business activity, such as in the middle of the night or during the weekend.
This may be true for those businesses that operate within certain days and times (i.e., retail), however, there are other business which need to operate around the clock (i.e., manufacturing) or have more unpredictable schedules (i.e., food processing). Especially for these companies, planned downtime can, in varying degrees, disrupt the operations, and the tools and processes that employees and customers rely on.
Additionally, not only does the extent of the impact of a planned downtime depend on a company’s operations schedule, but also on how the IT environment is set up – especially if it is centralized within one data center as opposed to being distributed across various regional locations across the country – and even the globe. IT systems are becoming more centralized due to the consolidation of organizational infrastructure as well as the increasing number of cloud migrations.
With the occurrence of these trends, if a company has only one instance of a critical business system that is necessary to running its operations, any outage, whether planned or unplanned, will have a negative impact, potentially leading to devastating consequences.
One example that illustrates how planned downtime can be just as costly as unplanned in terms of lost revenue measured per minute or hour comes from a client who is a Fortune 100 Oil & Gas company. It tasked us with reducing the planned downtime of its ERP systems during a TAR, or “turnaround event.” A TAR is required for plants and refineries in this industry to provide a planned period of regeneration where parts of its operations are turned off in order to complete inspections or conduct repairs. TARs can last anywhere from two weeks to two months, and can be very expensive to undertake, but they are necessary.
The downtime from the TAR can lead to millions of dollars in lost revenue per day, with these losses increasing as the outage is prolonged – due to diminished production and increased equipment and labor costs. So, to mitigate these losses, the company’s ERP system had to have the absolute lowest amount of downtime possible during this period so that it could effectively orchestrate the TAR process and allow the plant to go back to being operational more quickly.
2. The misconception that there is no way to control planned downtime – it is simply necessary for doing business.
Another reason planned downtime has historically not been given as much attention as unplanned is because IT leaders are used to thinking of it as necessary for doing business, and simply accept that it needs to happen, especially for legacy applications. However, with a modern approach and way of thinking, there are methods of controlling planned downtime instead of being forced to passively endure it.
Planned Downtime No Longer Has to be a “Necessary Evil”
The good news is that because of advances in technology, in many cases planned downtime is no longer the necessary evil it has been in the past.
Applications today are being built to run on clusters of servers using microservices architecture. This means that they generally can be updated and redeployed without requiring any downtime. These applications run multiple instances of the same services, allowing patches and updates to be made to one instance, while the rest of the instances are left operational. Once the modifications have been made to it, the updated instance can be switched back in real-time. Distributed, cloud-native architecture and orchestration solutions such as Kubernetes can take credit for systems needing less or no planned downtime for maintenance.
IT departments can apply the same update strategy to legacy apps that is possible for cloud-native apps, even when applications are not able to be refurbished to run as microservices on systems like Kubernetes, making planned downtime a thing of the past. As SAP began its life as an on-premises monolithic platform, it is necessary to implement the right automation and orchestration tools to both run cloud infrastructure configurations and capture specific application configurations.
On top of this, when pursuing immutable operations, it is important to consider that critical configuration information is located deep inside SAP databases, creating complexity for site reliability engineering (SRE) teams. However, when cloud automation is joined together with SAP configuration management, it enables the capability to develop parallel landscapes, apply upgrades and patches, and then fail forward to them with little to no disruption.
At Lemongrass we provide our clients with the benefit of our years of experience, knowledge and expertise regarding essential SAP configuration information, to realize near-zero downtime maintenance for SAP applications – which do not include the tooling needed to achieve it. This allows organizations to deploy upgrades and patches as they are needed, without requiring planned system outages. In this way, changes can be made quickly without interrupting business operations.
This will lead to a change in the attitudes of current IT teams and business leaders. They’ll come to see that planned downtime doesn’t have to be accepted as an unavoidable cost incurred to take advantage of new application features, upgraded functionality, security patches and/or bug fixes. Other benefits they will soon recognize include greater business agility and cost savings, since instead of multiple days, weeks, or months, the required system maintenance can be performed by only one administrator, and not teams of employees, in much less time.
Conclusion
No one can totally guarantee that unplanned outages won’t still occur. But because of enhancements in software design and technology, as well as changes in mindsets regarding no longer blindly accepting shutdowns as a necessary evil to achieving system enhancements and fixes, planned downtime is quickly becoming a thing of the past.
Routine maintenance no longer requires applications needing to go offline when companies utilize modern concepts such as automation and orchestration tools – whether they are using legacy apps, cloud-native applications, or a hybrid of the two.
Contact the SAP experts at Lemongrass to learn more about how we can reduce planned downtime by using our patent-pending MDOTM solution enabled by the Lemongrass Cloud Platform (LCP) technology.
