Optanix on AIOps – Part 6: Use Cases for AIOps
The following two examples of use cases for AIOps are based on the important concept of leveraging AIOps for a “time to threshold notification.” In an effort to better operationalize near-time capacity and availability constraints, Optanix sends notifications through a connector that feeds into the Optanix Platform.
Many of our clients are starting to adopt DevOps environments where the expectation is that they can turn up resources programmatically within hours, if not minutes, based on demand. Optanix has responded by provisioning an on-demand AIOps architecture.
With these resolutions, Optanix is able to generate workflow-based notifications that a given resource will be limited within X amount of time and to notify the provisioning platform to automatically spin up additional resources to immediately meet the resource demand.
Two Mini Case Studies on Use Cases for AIOps
Airline Use Case for AIOps
This airline’s IT contract was such that they paid monthly based on a high watermark of utilization. The expectation was that the carrier would manage need based on the airline’s customers’ demand. But here’s what happened:
The airline advertised a one-day sales promotion. Clients generally forecast an expectation of hopeful demand. However, in this case, the scheduled release date also happened to be a bad weather day. Web traffic soared as users were trying to find out status and rebook flights. At the same time, the sale drew unprecedented numbers. The carrier was unable to manage demand or meet SLAs.
Before Optanix, the demand crippled the infrastructure, making it completely unusable. The burden fell back on employees at terminals and in the contact center. The queue times and customer dissatisfaction rose and customers stopped calling in for the promotion. The sales promotion was an expensive loss.
With Optanix AIOps, however, the threshold alarms are received in context. Business events are correlated and additional Internet bandwidth is allocated to the necessary resources during promotion hours. Based on CPU, disk and memory performance, the Optanix Platform can identify two hours ahead of time that database systems will fail. And based on current traffic requests and past utilization baselines, the load balancer threshold determines when servers needed to be added to the web pool. Notifications are sent to the orchestrator, which provisions new services and resources to load balance and offset the database workload.
As a result, during promotions, the system averts degradation and failure and operates without interruption.
Healthcare Provider Use Case for AIOPS
This healthcare provider’s contract sanctioned Optanix to manage their resources and capacity. Here is the scenario:
The healthcare provider launched a new line of business and sent out a mass communication to its patients/members that these new services were available. The response to the request was off the charts. Call volume ramped rapidly. Simultaneously, the provider was in the process of replicating systems between data centers to add additional bandwidth. This increased workload was exacerbated by the system replications and predictably had a significant impact on performance.
Here’s where it went off the rails: three 10-Gbps connections between data centers normally ran at 30% utilization or lower. However, because the client was utilizing equal-cost multipath routing, the server move was based on source and destination. The connections were running around 95% utilization on one of the paths. This caused its normal 30% load to be pushed to the other two paths, increasing them to 45%.
The call volume then caused both of those circuits to crest above 50%. Monitoring alerted that there was over 80% utilization on a circuit, which was expected. No alert was generated for the other two circuits. Subsequently the system experienced performance degradation and failures.
Before Optanix, the client had a circuit outage and re-converged onto the other two circuits automatically. The traffic was too much and the Assured Forwarding queue for replication didn’t have a big enough buffer. The real-time Expedited queue couldn’t buffer the traffic well enough either. Call quality became unacceptable and images began corrupting.
With Optanix AIOps, threshold alarms are received in context to business processes. Now, when infrastructure issues related to bandwidth capacity occur, proactive measures can be automatically taken to, for example, stop replicating traffic. So, prior to failures, operations teams are notified and IT can divert the impending risk of resiliency loss.
The Uniqueness of the Optanix Platform: ITOM and MSP Benefits
The Optanix approach to AIOps adds value to managed service provider (MSP) and enterprise IT operations teams in several ways:
- It is focused on targeted use cases in the IT operations space, which makes it possible to realize AIOps benefits early in the deployment and adoption of the platform. It is not a generic big data tool – it is tailored to cater exclusively to IT operations requirements.
- It is extensible – the use of a data lake, the extensive automation libraries and tools, workflow automations, and microservices architecture make for rapid adoption of new data sources and fast deployment of features.
- By using a cloud-based management layer, the Optanix Platform facilitates multi-client access and streamlines remediation and workflow activities.
- It brings predictive focus to IT ops – using machine learning and smart analytics on a targeted set of monitoring data, the platform performs real-time validation and troubleshooting to provide insights as issues are occurring. This makes it possible to act before service degradation or outages set in, enabling a predictive, proactive and preventive approach to IT ops.
- It provides true business service assurance. The topology mapping of the infrastructure onto the application and service layers make it possible to understand the true impact to business services and enable prioritization of remediation issues.
- It is built by an MSP for MSPs. Based on decades of real-world experience, the Optanix Platform is purpose-built and provides proven IT ops capabilities.