User's Guide

Frequently Asked Questions 3

Event Monitoring Service Developer’s Kit

Programming Tips

How do I prevent initial monitoring requests from

timing out?

Using rm_monitor_start and rm_get_next_event

EMS monitors are launched in response to client requests. A client

request cannot be accepted by a monitor until two speciﬁc EMS API calls

are made. These two calls are rm_monitor_start() and

rm_get_next_event().

• rm_monitor_start() is the ﬁrst call made to the EMS API by a

monitor. It is called to register the process as a resource monitor, and

to establish a communication path between the EMS API and the

Registrar.

• rm_get_next_event() should be called next by the monitor so that

the monitor is ready to receive requests. It is through this call that

the client requests are passed to a monitor.

Typically, the monitor will block at the rm_get_next_event() until an

incoming message from the EMS API has been received. If the monitor

does not call rm_get_next_event() quickly enough, client applications

may timeout before a client’s request completes.

Figure 1-1 outlines the typical request process.