ems - Description of Event Monitoring Service (EMS)
is a monitoring service that polls a system resource and sends messages
when events occur. An is something you want to know about. For example,
you may want to be alerted when a disk goes down or when available
filesystem space falls below a certain level. EMS allows you to config‐
ure what you consider an event for any given system resource (disk,
filesystem space, network interfaces).
The advantage EMS has over built-in monitors is that requests can be
made to send events to a wide variety of software using multiple proto‐
cols (opcmsg, SNMP, TCP, UDP).
EMS consists of three components:
The framework starts and stops the monitors, stores information used by
the monitors, directs monitors where to send events. This framework
consists of the Registrar, the resource dictionary, and the EMS API.
A standard API provides a way to plug in new monitors as they become
available, or to write your own monitors; see the document "Writing
Monitors for the Event Monitoring Service (EMS)" available from the
high-availability web site: "http://www.hp.com/go/ha".
are applications written to gather and report information about spe‐
cific resources on the system. When you make a request to a monitor, it
polls the system information and sends a message to the framework,
which then interprets the data to determine if an event has occurred
and sends messages in the appropriate format.
A set of monitors are shipped with EMS: disk, cluster, network inter‐
face, and system resource monitors.
Runs under SAM (System Administration Manager).
EMS Directories and FilesEMS files live in and The following is a description of files and
directories that might help you as you navigate through EMS.
A file that sets the restart interval for monitor persistence.
A directory that contains resource dictionaries for the various moni‐
tors. The disk monitor resources are listed in and the cluster, net‐
work, and system resource monitors are in the If you were writing your
own monitor, the dictionary would go in this directory.
A directory where all the monitor daemons live. Some important daemons
in the directory:
restarts any failed monitors based on information in the file
handles passing monitoring requests to the correct monitors, and send‐
ing qualify events out in the correct protocol format.
A directory of log files used by EMS. The most interesting for trou‐
stores api calls made by monitors.
stores calls made by clients, such as MC/ServiceGuard or the SAM inter‐
face to EMS.
A command that lists the latest polled status of the specified resource
on a specified system.
AUTHOREMS was developed by Hewlett-Packard.
SEE ALSOdiskmond(1m), resls(1), emscli(1m)ems(5)