Performance and scalability
Watched Folder can serve 100 folders in total on one single
node. The performance of Watched Folder is dependent on the performance
of the LiveCycle server. For asynchronous invocation, performance
is more dependent on the system load and jobs that are in the Job
Watched Folder performance can be improved by adding nodes to
the cluster. Watched Folder jobs are distributed across the cluster
nodes by virtue of the Quartz scheduler and, in the case of asynchronous
requests, by the Job Manager service. All the jobs are persisted
in the database.
Watched Folder depends on the Scheduler service for scheduling,
unscheduling, and rescheduling the jobs. Other services, such as
Event Management service, User Manager service, and Email Provider
service, are available that share the Scheduler service thread pool.
This can affect Watched Folder performance. The Scheduler service
thread pool tuning will be needed when all the services start using