Oracle EBS Concurrent Managers Reference
Concurrent processing is the batch engine of EBS, and most operational problems trace back to a manager that is down, undersized or mis-specialized. This table explains each standard manager and the settings that tune it.
| Manager or setting | Type | Short name | Purpose |
|---|---|---|---|
Internal Concurrent Manager | Manager | FNDICM | Master process that starts, restarts and stops the other managers and watches their health. |
Standard Manager | Manager | STANDARD | Runs requests that are not assigned to a specialized manager. Usually the busiest queue. |
Conflict Resolution Manager | Manager | FNDCRM | Checks incompatibility rules and holds a request until conflicting programs finish. |
Scheduler/Prereleaser Manager | Manager | FNDSCH | Starts scheduled and on-hold-release requests at the requested time. |
Internal Monitor | Manager | FNDIMON | Monitors the Internal Concurrent Manager and restarts it if it fails (used in multi-node setups). |
Inventory Manager | Manager | INVMGR | Processes inventory transaction requests asynchronously. |
Receiving Transaction Manager | Manager | RCVOLTM | Processes receiving transactions entered online. |
Service Manager | Service | FNDSM | Service started on each node by Generic Service Management so the Internal Concurrent Manager can start processes remotely. |
Output Post Processor | Service | FNDCPOPP | Applies XML Publisher templates and post-processing to report output. |
Workflow Mailer Service | Service | Sends and receives Oracle Workflow notification email. Runs inside the Generic Service Component Container. | |
Workflow Agent Listener Service | Service | Dequeues and processes Workflow business events from agent queues. | |
Work Shift | Setting | Time window and number of target processes a manager runs. A manager can have several shifts per day. | |
Processes (target) | Setting | Number of requests a manager can run in parallel during a work shift. | |
Sleep Seconds | Setting | How long an idle manager waits before checking the queue again. Lower values give faster pickup at higher database load. | |
Cache Size | Setting | Number of requests a manager reads into memory per queue check. Often set to twice the number of target processes. | |
Specialization Rules | Setting | Include or exclude rules that decide which programs, types, users or applications a manager will run. | |
Incompatibility Rules | Setting | Programs that must not run at the same time, enforced by the Conflict Resolution Manager. | |
Request priority | Setting | Lower numbers run first when many requests wait. Default is 50. | |
adcmctl.sh | Script | Control script in the application tier scripts directory to start, stop or status the concurrent managers. | |
CONCSUB | Utility | Command line utility to submit a concurrent request from the operating system. |
Available managers depend on the modules licensed and installed. Short names shown are the standard seeded ones. Verify in Concurrent, Manager, Define on your instance. Never delete seeded managers.
Concurrent processing review
Share your manager definitions and get a short note on sizing, specialization and shift gaps.
Optional. The tool and every export work without this. No spam.
How to use the table
Search for the name you see in the Administer Concurrent Managers form or in a request log. Use the type filter to separate managers from services and tuning settings. When a request sits in Pending Normal, look at the Standard Manager and its work shift. When it sits in Pending Standby, a Conflict Resolution Manager rule is the usual cause.
- Managers run requests, services run background components
- Pending Standby points to incompatibility rules
- Inactive No Manager points to missing specialization or shifts
Sizing target processes, sleep and cache
Target processes limit how many requests run at once. Raising them helps only if the database and application server have capacity, so watch CPU first. Sleep seconds balances job pickup delay against polling load: a value of 30 seconds is common and 5 to 10 seconds is used for managers that must react quickly. Cache size should be at least the number of target processes so a manager is not starved while it waits on the next queue read.
Specialization in practice
A heavy program such as a month end cost run can be given its own manager with an include rule so it does not block short requests on the Standard Manager. Add an exclude rule on the Standard Manager only if you want that program to run nowhere else. Always test with a low-priority request and confirm the manager picks it up before relying on the rule.
Common mistakes
Setting target processes far above what the database can serve, leaving a manager with no work shift so it never starts, and bouncing the Internal Concurrent Manager during a long month end run are the usual errors. A normal stop waits for running requests, while a terminate stops them, so choose deliberately.
Frequently Asked Questions
What does the Internal Concurrent Manager do?
It is the master manager. It starts the other managers according to their work shifts, restarts them when they fail, and handles control commands such as deactivate and terminate. If it is down, no concurrent request will run.
Why is my request in Pending Standby?
Usually the Conflict Resolution Manager is holding it because an incompatible program is running. Check the Incompatibility Rules of the program, or the Conflict Resolution Manager itself if no incompatible request exists.
How do I start the concurrent managers from the command line?
Source the application environment file and run adcmctl.sh start with the apps username and password. The script is in the application tier scripts directory. Use adcmctl.sh stop and status for the matching actions.
Should I run more target processes to speed up batch?
Only when CPU, memory and database capacity allow it. More parallel requests on a saturated database slow every request. Measure load first, raise processes gradually, and check for locking between programs.
Netray can add an assistant that explains stuck requests and manager queues to your finance users in plain language, using read-only queries on your own instance.
Related Tools
Oracle EBS Profile Options Reference
A searchable lookup of the profile options EBS administrators change most: sign-on security, concurrent processing, multi-org, logging and session settings.
OracleOracle EBS Instance Health Assessment
A 10-question self-assessment that scores how well your EBS instance is run and points to the first areas to fix.
OracleOracle EBS Rapid Clone Checklist
Step-by-step checklist for cloning an EBS instance with Rapid Clone, including the post-clone scrubbing that protects production data and integrations.
OracleOracle EBS Inactive User Cleanup SQL Generator
Set a number of idle days and generate queries to list inactive users, expiring passwords, and a reviewed script to disable them.
More Oracle tools
See all- Oracle EBS Inventory Tables Reference
- Oracle EBS WIP Tables Reference
- Oracle EBS Cycle Count Workload Calculator
- Oracle EBS Discrete Manufacturing Go-Live Checklist
- Oracle EBS Inventory Accuracy Readiness Assessment
- Oracle EBS Concurrent Request Phase and Status Reference
- Oracle EBS Responsibility Design Worksheet
- Oracle EBS GL Period Close Checklist
- Oracle EBS AP Invoice Interface Template