Massively reduce the unscheduled downtime of SAP systems – in some cases by more than 70 percent: The first article in our series on IT automation in SAP operations shows how this can be achieved. We examine important basics, technical tricks, and use cases. We explain how SAP customers benefit from IT automation and what trends such as Artificial Intelligence Operations (AIops) are all about.
An Oracle database needs more storage space, a patch is required to close security gaps, and the update of an application has to reflect new compliance requirements: SAP landscapes never remain the same for long, but are changing continuously. Whereas such changes were previously documented in extensive runbooks and carried out manually by the IT team, automation tools are increasingly taking over these tasks.
The range of tools has changed over time. Initially, the focus was on incident and event management solutions. In other words, software that monitors all events in the infrastructure around the clock and automatically triggers a service ticket in the event of anomalies or malfunctions – or alerts the responsible administrator directly in critical situations. Smart tools meanwhile also take over configuration management or the orchestration of IT resources in a largely automated manner.
With AIops, artificial intelligence is now entering IT operations, raising application and infrastructure operations to a new level. AIops-based tools eliminate both faults and their causes independently, so that the same error does not occur a second time. At the same time, they establish causal relationships between different events in order to identify effects on operational processes at an early stage and to intervene directly if necessary. For example, if a large database query has a negative impact on performance, the AIops solution automatically allocates more computing capacity.
To implement predictive analytics in SAP operations, self-learning algorithms evaluate large amounts of data in real time. To give some idea of the sheer mass of data: T-Systems monitors more than 350 parameters every five minutes at application level alone. In addition, there is at least three times as much monitoring data from the network, server, and storage area. Around 17,000 values per SAP SID per hour and almost half a million checks per day.
To operate its customers' SAP systems efficiently and securely, T-Systems currently uses around 100,000 automated processes for simple and recurring tasks such as troubleshooting or database and operating system upgrades. In addition, there are more than 30,000 automated processes in which employees only have to take minor action.
The automation of SAP operations reduces manual effort and associated errors. This results in economies of scale that T-Systems passes on to its customers: Companies benefit from lower operating costs, less downtime, and reliable application operation. As a result, unscheduled system downtime is reduced massively, in some cases by more than 70 percent. At the same time, new features and SAP services become marketable more quickly. This is because development teams can also use the automation tools to shorten the design – code – deploy cycles in test systems.