A user best practice is to use a schedule to generate large documents, but not necessarily short ones. The administrator can configure synchronous job settings to prevent problems that can occur when users generate large documents unscheduled. The Factory generates unscheduled documents synchronously. A synchronous job is the Factory process that generates a temporary document.

Max synchronous job runtime limits the time a Factory can take to generate a document. Such a limit can prevent the generation of a huge unscheduled document from dominating Factory resources and degrading the response time of iHub. If an organization has mostly short documents, decreasing the value of the maximum execution time for on‑demand execution requests can improve performance. If an organization has a high ratio of system resources to users, increasing the value of the maximum execution time for on‑demand execution requests permits users to generate large, unscheduled documents.

Setting the value of Maximum execution time for on demand execution requests too high can delay generation of small, unscheduled documents, which can cause requests to back up in the queue and time out, and users to experience delays. Setting the value to 0 prevents the generation of any unscheduled documents.

Each synchronous resource group has a job queue. When Factories are busy, a request for a Factory service to generate a document waits in the queue. The administrator sets the number of requests that the queue can hold using Job queue size for synchronous reports.

If a user receives a message that the synchronous job queue is full, consider increasing the queue size or resources, such as CPU power. Setting the size of the queue too large can cause the accumulation of too many requests in the queue. A request can time out in the queue.

Job queue time-out for transient reports is the period of time, in seconds, that a request for an unscheduled document remains in the queue. Requests wait in the queue for 600 seconds, or 10 minutes, by default, then the request times out and fails. The user who requested the document receives a message that the job expired in the queue. The administrator can change the queue time-out by setting Job queue time‑out for transient reports.

Table 7‑2 lists the property names that appear in Configuration Console with the corresponding parameter names in acmetadescription.xml, indicating default settings, ranges, and when a property change takes effect.

By setting the Message Distribution Weight property, the administrator can assign priority to nodes for synchronous Factory service. By default, the number of Factories, compared to other nodes in the cluster, determines the weight of a node. The number of Factories in a node usually reflects the number of processors. The administrator can change the default weight of a node to take other factors, such as CPU speed, into consideration. The administrator can also change the weights of nodes if the synchronous Factories of a node are too idle or too busy.

The message distribution weight of a node is relative to the message distribution weights of other nodes. A node having a lower weight than other nodes receives fewer requests for unscheduled documents than the others. The message distribution weight of a node applies to all synchronous resource groups defined for the node.