Autonomous Troubleshoot: The New Oracle Monitoring and Advisory Service
-
Posted by Harry E Fowler
- Last updated 6/23/23
- Share
During Quest Forum Digital Event: JD Edwards Week, Claudiu Iosifescu, Senior Manager of Oracle Applications Support, gave an overview of Autonomous Troubleshoot – the Oracle Monitoring and Advisory Service for Applications Unlimited. This autonomous application checking service is designed to quickly report faults, proactively detect issues that might impact availability or a major business process, and improve your visibility and operational oversight of your Oracle JD Edwards EnterpriseOne environments. Iosifescu covered the service’s architecture and its automated ability to discover agents and targets utilizing an ever-growing library of hundreds of smart sensors and data collectors.
Have you ever wondered the following?
- How can we further automate the management of our application?
- How do we get ahead of impending issues and proactively detect them so we have time to address them?
- How can we focus our IT resources on advanced technology projects?
- How can I reduce the effort spent on handling faults during key business processes or daily operations?
The solution: Add smart sensors to your system!
Smart sensors autonomously scan known issue root causes, automatically create service requests with required collections, and route to senior-level JDE E1 experts.
Overview of the New Monitoring Service
Fast facts about the new Monitoring Service from Oracle include:
- Created for Oracle JDE EnterpriseOne 9.2
- Comes with a dedicated Technical Account Manager (the main point of contact for any questions you may have)
- Available for on-prem and cloud JDE customers
- Automatic platform and functional event detection uses Oracle Enterprise Manager—no other license needed
- Dedicated customer portal for this service
Here’s how a Smart Sensor Process Flow works:
- A fault occurs on a system monitored by JDE Service (automated by EM and gateway)
- Fault detected by the JDE plugin installed on OEM (automated by EM and gateway)
- Fault generates an event in the OEM Repository (automated by EM and gateway)
- The event sent by OEM agent and registered to Oracle Advanced Gateway DB (automated by EM and gateway)
- Event triggers MOS API on the gateway and automatically creates an SR (automated by EM and gateway)
- SR routed to the relevant JDE team and to the customer (automated by EM and gateway)
- Occurrence details tracked, KM docs attached, files/logs attached (automated by JD relevant team)
- The customer uses the SR to address the fault & implement a solution (customer interaction with JDE engineer)
- Customer confirms the fault is resolved and SR can be closed (customer interaction with JDE engineer)
Currently, there’s a large number of smart sensors available for each Oracle product:
- JDE – 240+ Smart Sensors
- PeopleSoft – 150+ Smart Sensors
- Siebel – 100 Smart Sensors
- E-Business Suite – 300+ Smart Sensors
Within the 240+ JDE smart sensors, there are 70 technical smart sensors and 170 functional smart sensors. Both technical and functional scanning happens with a variety of frequencies, depending on the root cause to be detected. The smart sensor database is continually increasing based on customer feedback and Oracle JDE Expert analysis.
To provide a picture of business impact, here is a description of six JD Edwards EnterpriseOne smart sensors.
Enterprise Server Processes Outstanding Requests
Description: This smart sensor will collect information on the number of outstanding requests held in each E1 kernel queue.
Business impact: When outstanding requests are kept in an E1 kernel message queue at high levels on a regular basis, that kernel performance will be impacted, which could translate into long-running transactions, time outs encountered by users, and eventually rollbacks of partially submitted transactions.
How it’s detected: The smart sensor detects the outstanding requests through an inquiry sent to Server Manager Management Console and this will reply back the number of the outstanding requests for each E1 kernel.
Enterprise Server Processes CPU Usage
Description: This smart sensor will collect information on the CPU usage of each E1 kernel.
Business impact: When high CPU usage is reported for an E1 kernel on a regular basis, this will impact the performance of the respective E1 kernel. This may lead to multiple effects, depending on the type of that E1 kernel. For example, for Call Object kernels, this may indicate a loop in the code caused by circular references in the Business Data or wrongly handled variables within the Business Function code.
How it’s detected: The smart sensor detects the CPU usage through an inquiry sent to Server Manager Management Console and this will reply back the CPU usage for each E1 kernel.
JAS Server JVM Heap Memory
Description: This smart sensor will collect heap memory values for each JAS Server JVM.
Business impact: Low levels of free heap memory, reported on a regular basis, indicate that Garbage Collector will be run very often. Since Garbage Collector is a “stop world” process, all the transactions could be delayed. This could also lead to stuck threads reported at the Weblogic Server level.
How it’s detected: The smart sensor detects the JVM free heap memory through an inquiry sent to Server Manager Management Console and this will reply back the memory value for each JAS instance registered in Server Manager Management Console.
Voucher Processing
Description: This smart sensor scans for Voucher batches that may be stuck during posting, and it helps users identify the list of Vouchers that were not fully processed.
Business impact: Partially processed Vouchers can lead to delayed payments to suppliers, causing a financial loss through the accrual of late charges or disqualified early payment discounts.
How it’s detected: The smart sensor scans the F0411 table and flags any voucher entry line that did not post successfully.
Failed Transactions and Accurate Stock Levels
Description: This smart sensor identifies failed inventory transactions where stock levels could be negatively impacted.
Business impact: Accurate inventory level is crucial to process efficiency. Incorrect stock level data can severely impact the efficiency of the supply chain, requiring manual intervention. Left unchecked, the build-up of incorrect stock records will amplify the level of effort needed to restore an accurate stock level.
How it’s detected: The smart sensor scans for records in the F41021WF table that are older than the pre-defined threshold.
Incorrect Currency Code on Account Balance
Description: This smart sensor scans for duplicate records in the Account Balances table.
Business Impact: Account Balances records are used by many programs and reports. Corrupted records may cause incorrect reporting and subsequent data corruption, for instance in the monetary account valuation report. Investigation of these issues can require significant resources and continuously growing efforts.
How it’s detected: The smart sensor scans the F0902 table and flags any account where the currency code field is both blank and not blank (e.g. USD, CAD, EUR) for the same fiscal year.
Benefits of Smart Sensors
Customers who utilize smart sensors receive several valuable benefits:
- Proactive improvement of system stability and availability
- Optimal system performance at all times and insights into CPU/memory trends
- Predictive fault prevention for critical business processes with recommendations for critical business events
- Local IT can be engaged in fixing issues before it impacts the business user community
- Cost savings through reduced admin tasks and higher business productivity
