Workflow Types (3.0)
Workflows are a collection of various types of agents to process data. The agents in a workflow are either collection agents, processing agents or forwarding agents. The list of agents that you can see, depends on your license for Usage Engine.
There are four kinds of workflows:
- Batch
- Realtime
- Task
- System Task
Batch Workflow
A batch workflow processes input that originates in a specific source, often a file. The workflow creates batches from the data and processes them one by one.
In a batch workflow data is collected by a single collecting agent in a transaction-safe manner: a single batch is collected (only once) and is fully processed by the workflow before the next batch is collected.
A batch workflow:
- Processes batches, one at a time
- Is started either manually or by a scheduled trigger
Stops either when it finishes processing the input, or when aborted
Note!
If the workflow aborts, the current batch can be reprocessed.
Real-Time Workflow
Real-time workflows are applicable in systems where instant processing requests need to be addressed as they occur.
In real-time workflows most of the collecting agents communicate in a two-way manner; they receive requests and provide replies.
A real-time workflow:
- Can have more than one collecting agent
Can process several UDRs simultaneously. See Multithreading(3.0).
Note!
Due to multithreading in real-time workflows, unlike batch workflows:
The order of UDR processing cannot be guaranteed.
Only global MIMs are used.
- Once started, is always active. A real-time workflow is started either manually or by a scheduled trigger, and stops either manually or due to an error.
- Processes in memory. Transaction safety must be handled prior to collection and after distribution.
- Real-time workflow error handling rarely leads to aborting the workflow. Errors are registered in System Log and the workflow continues to run.
Note!
Real-time workflows may use the Inter Workflow or Kafka agents to forward data to a batch workflow.
Task Workflows
Task workflows are single agent workflows that execute user defined SQL, shell scripts or unit tests.
System Task Workflows
A number of System Task workflows are included in Usage Engine. System Task workflows and workflow groups that include System Task workflows enable you to perform background routines such as log- and run-time data clean up from the platform.
System Task workflows include:
Archive Cleaner
Configuration Cleaner
Data Veracity Maintenance
- ECS Maintenance
Statistics Cleaner
System Backup
System Log Cleaner
- User Cleanup
This section includes information about:
Opening a System Task Workflow
- Modifying a System Task Workflow Configuration
Opening a System Task Workflow
You open a System Task Workflow from the workflow configuration.
To open a System Task, double-click a System Task workflow or workflow group in the Configuration Browser pane.
Modifying a System Task Workflow Configuration
You modify all the System Task Workflow configurations at template level. The workflow properties are all set to Final and cannot be modified.
- You can modify a System Task configuration, including its scheduling criteria, but you cannot create or remove a System Task configuration.
- The Archive Cleaner Workflow lets you modify only its scheduling criteria. For further information see Scheduling in Managing a Workflow Group(3.0) .
To Modify a System Task Workflow:
- Open the System Task Workflow that you want to modify.
System tasks list - Double-click the agent icon.
The agent's configuration dialog opens. - Perform the changes and click .
- Save the workflow.
Archive Cleaner
The Archive Cleaner System Task enables you to remove old archived files from the file system. Archive Cleaner operates according to data that it receives from the Archive profiles.
Note!
You can modify only the scheduling criteria of the Archive Cleaner. See Archive Profile. Since scheduling can only be applied to workflow groups, you modify the Archive Cleaner scheduling from the workflow group configuration.
The Archive Cleaner System Task placed in a workflow
Configuration Cleaner
Enables you to specify the maximum age of an old configuration before it is removed.
Note
You cannot remove the most recent configuration with the Configuration Cleaner, only historical ones.
To configure the Configuration Cleaner:
- Select a configuration type from the table and then click the entry in the Keep column; a drop-down list appears.
- Select one of the following: Always, Days, or Versions.
The Configuration Cleaner System Task
Item | Description |
---|---|
Type | The icon representation of the configuration type. |
Name | The name of the configuration type. |
Keep | Select the time unit with which you specify how long old configurations should remain in the system.
|
Value | Specifies the number of days or versions that represent the period during which configurations are kept. |
Data Veracity Maintenance
The Data Veracity Maintenance Task
The Data Veracity Maintenance system task removes outdated Data Veracity data, provided that the state is REPROCESSED or DELETE_APPROVED. For more information about the Data Veracity Maintenance System Task workflow see Data Veracity Maintenance System Task(3.0).
ECS Maintenance
The ECS Maintenance System Task
Enables you to remove old ECS data from the file system, provided that the state is REPROCESSED. For information about the ECS_Maintenance System Task workflow see ECS Maintenance System Task(3.0).
Statistics Cleaner
Enables you to remove old statistics data that has been collected by the Statistics server and stored in the database.
The Statistics Cleaner System Task
Item | Description |
---|---|
Minute Level Records | Specifies the number of days during which a minute-level record should be kept in the database. |
Hour Level Records | Specifies the number of days during which an hour-level record should be kept in the database. |
Day Level Records | Specifies the number of days during which a day-level record should be kept in the database. |
System Backup
Enables you to create a backup of all the configurations. A backup file is saved on the host machine where the platform application is installed.
The System Backup files are stored under $MZ_HOME/backup/yyyy_MM
, where yyyy_MM is translated to the current year and month and $MZ_HOME/backup
is the default path. The system saves a backup file and names it according to the following format: backup_<date>.zip
. System Backup also enables you to specify the maximum age of backup files before they are removed from the host disk. You can change this location using the following command:
<property name="mz.backup.path" value="${mz.home}/backup"/>
Refer to Persistent Storage (3.0) to know about how you can configure persistent storage.
The System Backup configuration dialog box is made up of two tabs:
System Backup: Lets you enable the backup option
Cleanup: Lets you configure the time during which a backup should be kept on disk before it is deleted
The System Backup System Task workflow - the System Backup tab
Item | Description |
---|---|
Enable System Backup | Select this check box to enable the system backup. The default value is On. |
Use Encryption | Select this check box to enable encryption of the backup. |
Password | Enter a password. |
The System Backup Cleaner System Task - the Cleanup tab
Item | Description |
---|---|
Imported Files | Every time the System Importer imports a configuration to the system, MediationZone saves it as a backup on the Platform. Enter the period, in days, during which the imported files should remain on disk. |
System Backup Files | Defines the maximum age of system backup files before they are removed from the host disk. |
System Log Cleaner
The System Log Cleaner deletes the System Log periodically. You set the frequency values for deleting different message types, on the System Log Cleaner dialog box.
The System Log Cleaner System Task
Item | Description |
---|---|
Error/Disaster | Enter the maximum age of Error and Disaster messages before they are removed from the database. Max value: 99 days Default value: 30 days |
Warning | Enter the number of days during which Warning messages should be kept in database. Max value: 99 days Default value: 20 days |
Information | Enter the number of days during which Information messages should be kept in the database. Max value: 99 days Default value: 10 days |
User Cleanup Task
The User Cleanup Task can be used for automatically deleting users who are no longer active. If SSO is turned on, any user logged in through SSO will be added upon login, but the user will not be removed when it is removed from the OIDC server. Because of this, the User Cleanup Task can be used for cleaning out any user that has not logged in for a configurable number of days.
The User Cleanup Task
Item | Description |
---|---|
Apply only to SSO users | This is checked by default and means that only users created through SSO login will be deleted by the task. |
Keep Days | The default is 0. The number of days to wait before deleting a user after the last time they logged in. A value of "0" will not delete any users. |
Default Successor | Normally a user cannot be deleted if the user stands as the owner of any configuration. Because of this, you can choose a "successor" which is a user that will take over the ownership of any configuration when the first user is deleted. Default successor is meant for choosing a user to use as a successor if a user that is to be deleted doesn't have any successor configured. |