The workflow template is the area where the workflow is designed. The workflow template outline one or several workflows.

Workflow configuration

You create a workflow configuration by using one of the following methods:

• Click the New Configuration button in the upper left part of the Desktop window, and then select  Workflow from the menu. Select a Workflow Type : Batch, Realtime, or Task, and then Create.

• Drag agents from the agents pane onto the template.

• Double-click an agent on the agent pane.

• Select an agent from the Agents menu.

When the first agent is placed on the workflow template, one row is automatically created in the workflow table.

To create a data flow, agents need to be connected to each other. To connect the agents, click the left mouse button on the center of the source agent and without releasing the left mouse button, move the pointer to the target agent and release there. This creates a connection (route) between the two agents indicating the data flow.

All editing and triggering from the workflow template generate changes to the workflow configuration through the workflow configuration. Examples of this are adding and removing agents, altering agent positions and editing agent settings and preferences. The workflow table will be affected if it includes columns that correspond to an agent removed from the workflow template.

When an agent is deployed into the workflow template it receives a default name located underneath it. The same applies to routes when they are added. These names may be modified to ease identification in monitoring facilities and logs.

Resting the cursor over an agent in the template, displays parts of its configuration in a tool-tip text.

An agent tool-tip of a workflow configuration

Configuration

Each agent in the workflow configuration has a specific configuration dialog named after the agent type. Each route in the workflow is by default either asynchronous or synchronous, but this can also be configured per route. You can access these configurations by double-clicking the agent or route, or selecting the agent or route and right-clicking to reveal a pop-up menu. The pop-up menu contains different options depending on if you have selected an agent or route.

Right-click menu for agents

The right-click menu for agents contains the options Configuration... , Cut/Copy/Paste/, MIM Browser... , and Workflow Properties .

Right-click menu for agents

The configuration dialog contains configuration information located in tabs. The leftmost tab contains configuration parameters that are unique for each agent while possible additional tabs are of a more generic kind and may be recognized in other agent windows. A description of components in the first tab is available in the user documentation section for the respective agent, while the remaining tabs are described in the section below, Agent Services.

Agents and their configurations can be copied using the copy/paste functions in the Edit menu. Select the source agent followed by Copy and then Paste. This will deploy a copy of the selected agent and its configuration.

You can modify an agent name by selecting the name (clicking on it) and typing a new name. Agent names can also be edited in the configuration dialog, displayed when the agent is double-clicked. Agent names must be unique within a workflow configuration and may only contain the a-z, A-Z, 0-9, "-" and "_" characters.

Right-click Menu for Routes

The right-click menu for routes contains the options Configuration, Copy/Paste/Cut, MIM Browser, Workflow Properties, and Route Styles.

Right-Click Menu for Routes

When you select Configuration, the Route dialog opens up.

Route dialog

If you do not want to use the default configuration for the selected route, select the Override Default check box and select either of the Asynchronous or Synchronous options, and click OK to save your changes. A small A (for Asynchronous routes) or S (for Synchronous routes) will be visible at the start of the route. If you want to see the type for all routes, and not just for the ones you have stated explicitly, you can select the option Show All Route Types in the Preferences dialog, see the section below, Visualization, for further information.

When the route is synchronous, the same thread continues to process the UDR in subsequent agents before the route operation returns in order to process the next UDR. If the route is asynchronous, there is a queue associated with the route. Routing places the UDR in the queue for future processing, and the route operation returns immediately for the next UDR. 

In an Analysis or Aggregation agent, the route operation is done through the udrRoute command. For a synchronous route, which is the default for the Analysis agent, udrRoute can take a significant amount of time to finish since it waits until all synchronous processing in subsequent agents in the workflow is complete. If the route is asynchronous, which is the default for the Aggregation agent, the call to udrRoute returns immediately.

UDRs that are left on queues are picked up and processed by the workflow threads. The number of workflow threads can be configured in the Execution tab of the Workflow Properties dialog for the workflow.

If you click on the Route Styles option, you can determine the appearance of the route; Orthogonal, Bezier, or Straight.

A route name is modified by selecting the name (clicking on it) and typing a new name. Route names must be unique within a workflow configuration and may only contain the a-z, A-Z, 0-9, "-" and "_" characters.

Agent Services

Some agent configuration dialogs are optionally equipped with additional tabs holding configurations for different common services. This section describes the general services that are available in . The user guide for each agent, will in turn describe what services it uses.

Thread Buffer Tab

By default, a batch workflow utilizes one active thread at a time. By configuring a buffer storage for an agent, it will be possible for yet another thread to be created, this is also called multithreading. One thread will be populating the buffer, and another pulling it for data. Adding yet another buffer for another agent will add yet another thread, and so on.

This is especially useful in complex workflows with many agents. All batch agents that receive UDRs can utilize this functionality.

Open the Configuration dialog of the agent and select the Thread Buffer tab. The tab is present in batch processing and batch forwarding agents.

The Thread Buffer tab

A UDR may be queued up while another thread is busy processing a reference to it. Workflows routing the same UDR on several routes and involving further processing of its data, must consequently be reconfigured to avoid this. A simple workaround is to route the UDR to an Analysis agent for cloning before routing it to the other agents (one unique clone per route). See discussion under 9.6.4 Analysis Agent Assignment and Cloning about the potential problems with multiple agents processing the same UDR in parallel.

Analyzing Thread Buffer Statistics

By using the Print Statistics alternative in the Thread Buffer Tab , buffer statistics are logged for the whole batch execution and show the full and empty percentage for the threads utilizing the thread buffer. For information about multithreading in a batch workflow, refer to the section, Threads in a Batch Workflow, in 3.1.2 Multithreading.

11:03:04: Buffer usage statistics [5] for 2089 turnover(s) of UDRs: 
Incoming queue: Available 54%. Full 46%. 
Outgoing queue: Available 59%. Empty 41%.

The percentage values for Empty and Full must be as low as possible, and as equal as possible. The latter may be hard to achieve, since the agents may differ too much in processing complexity. If possible, add and configure another agent to take over some of the processing steps from the most complex agent.

See the section below, Thread Buffer Tab, for how to configure the thread buffer.

Filename Sequence Tab

For batch collection agents such as Disk, FTP, and SFTP, there is a service available, found in the agents' configuration dialog in the Filename Sequence tab. The filename sequence is used when you want to collect files that contains a sequence number in the file name. The sequence number is expected to be found on a specific position in the file name and has a fixed or dynamic size.

Filename Sequence tab

Sort Order Tab

The Sort Order service is available on some batch collection agents and is used to sort matched files before collection.

The sort pattern is expected to occur on a specific position in the file name or to be located using a regular expression. However, if the sort pattern cannot be applied, the workflow will abort, except for files with a name which is shorter than Position plus Length. For information on this exception, see the description below for Position.

Sort Order tab

 FILEA_1354.log
 FILEB_23.log
 FILEC_1254.log,

FILEC_1254.log
FILEA_1354.log
FILEB_23.log

Filename Template Tab

The Filename Template service is available to batch forwarding agents that are responsible for creating a file, and the Disk forwarding agent for real-time workflows. The configuration contains MIM resources for all available agents in the workflow, whose values may be used when constructing a filename for the outgoing file.

Since this service includes a selection of MIM resources from available agents in the workflow, it is advised to add all agents to the workflow, and to assign route and agent names, before the filename template configuration is completed.

Filename Template tab

The list contains MIM resources or user defined values that will create the file name. The order of the items in the list controls the order in the file name.

There are two ways in which you can determine the order of the items. You can use the Up and Down arrow buttons, or you can use the column headers. If you begin to sort the items using the column headers, the Up and Down arrows are disabled.

The table containing MIM resources, user defined values, separators and/or directory delimiters will create the filepath or filename. The order of the items in the table define the filepath or filename.

Since the service utilizes a selection of MIM resources from available agents in the workflow, it is advised to add all agents to the workflow before the filename template configuration is completed.

Filename Template tab - Add or Edit dialog

DO1-FILE1-000100.DAT
DO2-FILE1-000250.DAT
DO1-FILE2-000100.DAT
DO2-FILE2-000250.DAT

Defining a MIM Resource of FNTUDR Type

A MIM resource with a value of the FNTUDR type included in the filename template is treated somewhat differently than other MIM resources. A FNTUDR value is a text string that can contain delimiters. The delimiters in the FNTUDR value will be replaced by directory delimiters when determining the target file path. The FNTUDR is defined in the FNT folder.

import ultra.FNT;

    mimPublish( global, "FNTUDR Value" , any );

consume {

    FNTUDR fntudr = udrCreate(FNTUDR);

    fntAddString(fntudr, "dir");
    fntAddString(fntudr, "1");
    fntAddDirDelimiter(fntudr);
    fntAddString(fntudr, "dir2");
    fntAddDirDelimiter(fntudr);
    fntAddString(fntudr, "partOfFileName");

    mimSet ("FNTUDR Value", fntudr);

    udrRoute(input);

}

dir0/dir1/dir2/partOfFileName20070101_1

For further information about how to manipulate FNTUDRs with APL functions and how to publish MIM resources, see the APL Reference Guide and the section, Meta Information Model, in 1.2 Administration and Management.

Visualization

In the workflow configuration, zoom in or out the workflow illustration by modifying the zoom percentage number that you find on the tool bar. The default value is 100(%). To change the zoom value click the increase or decrease icons. Clicking the button between these icons will reset the zoom level to the default value. Changing the view scale does not affect the configuration.

It is also possible to change the appearance of the workflow template in other ways. You can open the Preferences dialog found in the Edit menu.

The Preferences dialog