Message Server Overview, Message Server Configuration, Message Handlers, Forwarders
Each message handler and forwarder configured for a message server can be assigned a retry schedule. The purpose of a retry schedule is to maintain persistent connectivity among machines on a network despite network failures or machine reboots. Retry schedules are configured through the Retry Schedule Dialog, accessible from the Configurations Dialog.
A retry schedule consists of one or more sequences. Each sequence specifies a number of attempts, an interval between attempts, and a logging option. The last sequence is special in that the number of attempts may be infinite. For each sequence, the developer can specify whether an internal notification message is generated in the event of failure at each attempt, once at the end of the sequence, or never. For example, the developer may specify that a forwarding connection be attempted 6 times every 5 minutes with no logging, then 5 more times every 30 minutes with a log message generated after the fifth attempt, and finally once a day thereafter, with a log message generated for each failure.
Each time a message handler is successfully loaded, a connection to a forwarder is established, or the message server is cycled, the retry schedule will be reset. This ensures that the system will not run out of retries; the schedule specifies how to respond to each failure.
An internal notification will always be generated if a message handler or forwarder fails and will not be retried. This can happen if an unrecoverable error occurs (e.g. the message handler's Object ID is not registered), or if the end of the retry schedule is reached.