Event Manager configuration
A single configuration file contains parameters that govern various aspects of the Event Manager behavior. The Event Manager supports event notification from the MDM operational server to external systems.
com.ibm.mdm.mds.event.manager.cfg
The configuration file is available at <WAS_INSTALL_HOME>/AppServer/profiles/<YOUR_APP_SERVER>/InstalledApps/<CELL_NAME>/MDM-native-<INSTANCE_ID>.ear/native.war/conf/com.ibm.mdm.mds.event.manager.cfg.Parameter | Description | Default and examples |
---|---|---|
enabled | Whether this operational server instance uses an Event Manager. | true |
threadCount | The number of contexts reserved for the Event Manager component. This controls
the number of parallel requests that can be submitted by the Event Manager to the MDM operational
server. Important: Do not change the default value (
1 ). Doing so will
negatively impact system performance. |
1 |
entityType | A comma-separated list of entity types to be managed by the Event Manager. The
default, *, manages all entities. In most cases, you should not need to change this parameter. |
* |
pollSeconds | If the previous attempt to select rows from the input queue returned no
results, pollSeconds defines the number of seconds to wait before the request
is reattempted. In most cases, you should not need to change this parameter. |
10 |
purgeSeconds | The number of seconds the purging thread waits if there are no completed
records to process. In most cases, you should not need to change this parameter. |
0.5 |
workUnit | The number of rows to select from the input queue (entoque record) to be
worked. In most cases, you should not need to change this parameter. |
500 |
workerThreads | The number of worker threads that are used to process work items (notifications) in parallel. | 1 |
maxQueueDepth | The maximum queue depth to be assigned to a queue manager while waiting to be
processed. In most cases, you should not need to change this parameter. |
5000 |
inputWrkOwner | The input work owner name. Work is selected from the input queue that is
assigned to this work owner. This parameter must not be the same as the
queueWrkOwner or errorWrkOwner properties. In most cases, you should not need to change this parameter. |
mdmadmin |
outputWrkOwner | The output work owner name. Work is returned to the queue assigned to this
work owner. This parameter must not be the same as the queueWrkOwner or
errorWrkOwner properties.In most cases, you should not need to change this parameter. |
(blank) |
queueWrkOwner | The queue work owner name. This setting is the work owner that is assigned to
an item while it is being processed. If you have multiple servers running against the same database, each server must have a unique queueWrkOwner. Note: If you are running in a clustered
environment, you must set unique queueWrkOwner properties for both the Event Manager and the entity
manager.
|
queue |
queueType | The type of queue that is being managed. Possible values are
input and output. In most cases, you should not need to change this parameter. |
output |
errorWrkOwner | The work owner name that is assigned if there an error occurred when this item
was processed. This property can be set to a custom string (for example,
errorWrkOwner=systemerrors ). When unrecoverable errors are encountered, the queue
work owner is set to the error work owner. Only applicable if
errorDequeueEnabled=true . If errorDequeueEnabled=false and the
maximum number of retry attempts are reached, then the record is removed from the database. In most cases, you should not need to change this parameter. |
error |
errorMaxRetryWaitSeconds | The number of seconds that the queue manager waits before it retries a
database operation that exceeds the errorRetryAttempts value. This value comes into
play only on queue manager operations that involve an mpi_entoque_xx that cannot be de-queued or
ignored. The minimum value for this parameter is 30 seconds.In most cases, you should not need to change this parameter. |
60 |
errorRetryWaitSeconds | The number of seconds to wait in between retry attempts. In most cases, you should not need to change this parameter. |
5 |
errorRetryAttempts | The number of attempts at retrying an item of work before one of these actions
are taken:
In most cases, you should not need to change this parameter. |
10 |
errorDequeueEnabled | Enables the de-queue of failed work items. When set to true ,
records are de-queued. "De-queue" means that when the number of retry attempts are met, the wrkOwner
of the queue record is set to errorWrkOwner . When the records are ready for further
processing, you must manually reset the queue record wrkOwner back to 'system'. When set to
false , the records are removed from the queue after the specified number of
attempts are reached (as defined in the errorRetryAttempts property).In most cases, you should not need to change this parameter. |
true |
enableAutoSuspension | When this property is set to true and an unrecoverable error
is encountered by the Event Manager, then all event notification is suspended. If you have multiple
Event Managers that are configured, processing stops for all of them. |
|
inclFullCVOnEIDChange | This system-wide property overrides the Send updates
only setting in InfoSphere®
MDM Workbench. If
set to true , entity events that are triggered by an Entity ID change will contain
all of the composite view attributes in the event message. If an event does not move a member to
a new entity, then the Event Manager accepts the Send updates only setting. A
|
|
workPriorityEnabled | Enables the processing of work items by priority order. In most cases, you should not need to change this parameter. |
true |
maxWorkPriority | Sets the maximum entity priority to be processed by this queue manager. The
higher the number, the lower the work priority. Only applicable if
workPriorityEnabled is set to true. In most cases, you should not need to change this parameter. |
0 |
workManagerName | The JNDI service name that this background manager runs under. The full
service name would be osgi:service/wm/workManagerName. In most cases, you should not need to change this parameter. |
default |
workManagerUserName | The user that runs the queue manager background threads. In most cases, you should not need to change this parameter. |
(blank) |
workManagerPassword | The password for the user. In most cases, you should not need to change this parameter. |
(blank) |
workManagerRealm | The authentication realm that the workManagerUserName is
authenticated against. In most cases, you should not need to change this parameter. |
defaultWIMFileBasedRealm |
enableEntityEventStaging | When this property is set to true , the Event Manager places
all outgoing entity events into a staging table in the MDM database (mpi_evtstager). The
true setting ensures that only one outgoing entity event for a given entrecno and
enttype is broadcast at a single time to each of its registered destinations. When the property
is set to Note: When entity event staging is enabled for virtual MDM, all event destinations (events
that are registered in the virtual model) must be configured to subscribe to either member events or
entity events only. You cannot subscribe to both event types.
|
false |
stagingWaitSeconds | The time, in seconds, that the Event Manager will periodically wait during staging before processing the duplicate entity events in mpi_evtstager table. | Default setting is 0.5 |