Following on from my previous blog on scale improvements in AIOps 4.9.1, I am happy to share the latest scale numbers.
First, a comment on how we arrive at these scale numbers. Our numbers are the result of rigorous testing that takes into consideration how customers use the system: The sources they connect to, how large the Alerts are, and how often the data is updated. To be able to support and stand over a number we must be confident that all customers will be successful at the scale point. It is important to note that our documentation has a few notes which lists details of our testing and assumptions.
Scale is something that we will continually improve upon as customers are looking to Unify more data into AIOps from their disparate systems.
AIOps now supports:
- 800,000 standing Alerts. This is double from my previous blog in 4.9.1 which also announced a doubling of our standing Alert count. Why is this so important? Standing Alerts represent active issues. Active issues are highly dynamic as they are worked on and updated via automation, the system that generated them, and/or users. Active issues are also grouped according to Temporal, Topological or Scope Based grouping, and have many user and system policies act upon them to suppress, promote, enrich, or take an additional action.
- 300 concurrent users. More users acting upon more data. Tackling the increase in users in parallel with increasing data shows our system was engineered and designed well to scale.
- 25 million topological resources - an increase of 25% over the previous limit. Our topology capability allows customers to stitch together topologies from many systems to provide a unified view of their data. You can read more in this blog: https://community.ibm.com/community/user/blogs/matthew-duggan/2026/01/07/aiops-4120-topology-scale
- six million metrics every five minutes - an increase of 20%. Metrics are prolific and abundant. For customers that choose to have AIOps analyse their metric data they can now analyse more than ever.
All of these increases - Alerts, Topology, Metrics, concurrent users, reflect our commitment to supporting our largest customers and those with large amounts of data unify their operations within IBM CloudPak for AIOps.