Introduction
As digital transformation accelerates and IT environments become increasingly complex, organizations need reliable, real-time visibility across their technology stacks. That’s where application performance monitoring and end-to-end observability can help. Monitoring is the process of tracking and collecting key performance metrics to pinpoint potential problems that may arise and fix them before they become an incident.

Today’s world enterprise wants to focus more on business than supporting their IT systems. As part of this many Enterprises today wants to adopt cloud journey and thereby reduce total cost of ownership [TCO]. There are multiple players in the cloud space who offer many services, Tool adoption, latest technology framework. Cloud offering comes in different XaaS[anything as a service] flavors like Infrastructure as a Service (IaaS), Platform as a Service (PaaS), Software as a Service (SaaS), Function as a Service (FaaS), Data as a Service (DaaS), Database as a service (DBaaS) etc.

Oracle being the major cloud service provider, they offer both an integrated suite of cloud applications and the cloud infrastructure platform. Oracle provides a complete set of mission-critical retail solutions under SaaS subscription. It is important to note that customer wants to make sure their business is not disrupted. It is apparent that all cloud providers offer sophisticated monitoring solutions either as service or with the leading market tools to monitor applications, middleware, integration components and underlying infrastructure. Likewise, RMFCS – A cloud SaaS subscription for Oracle Retail Merchandising Foundation Cloud Service has its own monitoring tool, however customer does not have any visibility or control over these performance statistics.

Based on this, there was a strong desire to develop a monitoring solution for a cloud platform. Infosys has developed a monitoring solution for a cloud platform that caters to customer requirement to monitor entire eco system including SaaS Infrastructure. This helps to get real-time data, alerts, and notifications and to identify and mitigate any of the future failures or attacks that may pose potential disruptions to operations.

Building Blocks
In general, all components of a platform are typically Monitored from 5 perspectives as explained below.
1. Application Module Monitoring
This includes code profiling, Thread monitoring, Heap utilization monitoring.
2. End User Experience Monitoring
This comprises of Synthetic transaction monitoring, Active user Monitoring etc.
3. Application Interface Monitoring
Any API calls or LDAP or connection to any source will be monitored here.
4. Middleware, Integrations & DB Monitoring
Any integration tools or services or Batch jobs or Database performance are monitored.
5. Infrastructure Monitoring
This is very vast and critical that includes all major infrastructure components to be monitored such as Cloud VMs or Network etc.

Few of these perspectives are adopted to Oracle Retail Merchandising Foundation Cloud Service as well.
In this context, there are few metrices identified as crucial to be monitored in Oracle Retail. This article explains on those important KPIs or metrices and how they are monitored using various market led tools and AI/ML solution by Infosys.

• Business Process Monitoring
• Data Replication, GG Service Availability
• Job/Batch Monitoring
• Application Interfaces Monitoring
• DAS – Infrastructure monitoring

Business Process Monitoring
Reports are created/built for all business-critical transactions such as PO, Price Changes, Transfers, Allocations, Invoices, Sales, Inventory, SL. This provides insight into any discrepancies across systems/applications.

Data Replication, GG Service Availability:
In Oracle Retail, there is no access to Data on SaaS unless they are replicated to extension layer that is called DAS or Extended Integration layer. Golden gate is used to replicate SaaS Data onto Extended Integration layer Data Base either on IaaS or PaaS. Monitoring GG service is critical to ensure data integrity across all connected systems. GG service is monitored through heartbeat and fed to AIOps engine to create an incident in case of any upcoming issues such as slowness or services not available etc.

Job/Batch Monitoring
Nightly batch job execution is very critical and is the prominent event in the Retail to process the data for the day and make systems ready for the next business day. It is utmost important to monitor these job executions for any failures. POM is used to schedule all batch jobs under SaaS platform. POM provides REST APIs to access status of each job. In retail there are multiple custom jobs that are scheduled separately outside of SaaS platform. REST APIs are leveraged to integrate these extended jobs with POM scheduled jobs. All these jobs are monitored through REST API to track status of each batch execution in call back mechanism. Metrics are mined through robotic process automation[RPA].

Application Interfaces Monitoring
Integration Services include monitoring of Integration components, API Calls, FTP Services, Message Flows. Data synchronization across systems/applications are crucial and ensured by monitoring all these integration services. These vital parameters are monitored and fed to AIOps engine for further process.

DAS – Infrastructure monitoring
DAS is an extension layer for Oracle Retail SaaS. Entire DAS infrastructure is hosted under customer platform be it IaaS or PaaS or On Prem, customer will own the responsibility of monitoring all resource starting from server/VM till the services. Above diagram shows what parameters are monitored on Oracle Retail Applications and how. In the broader perspective, the entire eco system is leveraging multiple market driven tools with combination of Infosys developed solutions.

Monitoring Tool set for entire eco systems:

Below diagram shows how all alerts are classified and fed to bot factory for further processing. This is entirely automated process where alerts are received based on the anomalies within the system and takes corrective action. These corrective actions are automated based on the automated model retraining [MLOps] over the period.

Cognitive-First Framework by Infosys for Business-Aware Operations

In conclusion
It is very important to monitor entire eco system including SaaS to detect any system anomalies to reduce the risk of business disruption, avoid organization’s services becoming unavailable and to enable customer to take appropriate action well in advance. Issues classified based on model learning, are bifurcated to right stream to take corrective actions. Unclassified issues undergo model learning though automated process [MLOps]. Corrections are possible before any abnormalities become an incident. So, prevent abnormal situations in the first place through sophisticated monitoring solutions covering entire customer landscape including SaaS, PaaS, IaaS, and On-Premises systems/Applications.

Automated monitoring & corrective solution by Infosys is the unique solution that caters to retailers need and can be leveraged by all retailers to enhance their business operations run smoothly.