Hardware Monitoring

This section addresses the following topics:

Event Monitoring Service

The Event Monitoring Service (EMS) is a framework that supports hardware monitoring. The EMS framework supports the EMS Hardware Monitors, the EMS High-Availability (HA) Monitors, third party monitors, and so on. Using EMS, you can manage the Peripheral Status Monitor (PSM) and EMS HA Monitor requests.

The EMS framework provides the necessary notification methods such as sending e-mail notifications to the root user and communicating with HP OpenView.

EMS Hardware Monitors

EMS Hardware Monitors are daemons that proactively monitor hardware devices such as CPU, memory, and hard disks. These monitors work with the EMS framework to detect and report hardware problems. When a monitor detects an error or an abnormal behavior, it generates an event that includes details such as a summary of the event, the cause of the error, and the recommended action, and it reports them to the EMS. EMS uses the monitoring request to determine whether and how to deliver the event.

EMS Hardware Monitors require minimal maintenance once installed and configured. They automatically monitor new hardware resources that are added to the system.

The EMS Hardware Monitors also integrate with applications that maintain system availability, such as HP Serviceguard.

Startup Client

The Startup Client launches and configures the EMS Hardware Monitors each time you start the system and after you execute the IOSCAN utility. It uses the default monitoring requests or any customized requests created to start each monitor and configure its hardware resources.

Hardware Monitoring Request Manager

The Hardware Monitoring Request Manager (monconfig) utility enables you to create and manage event monitoring requests for all the EMS Hardware Monitors. It supports all the notification methods. Using the monconfig utility, you can also enable or disable hardware monitoring.

The EMS uses the settings in the monconfig utility to determine how an event must be delivered. For more information about the monconfig utility, see “Hardware Monitoring Request Manager”.

Multiple-View and Non-Multiple-View Monitors

EMS Hardware Monitors are of two types: Multiple-View and Non-Multiple-View monitors. Multiple-View monitors enable you to specify different event details for the same monitor to one or more targets, based on the requirements of the targets. Event details for Non-Multiple-View monitors are the same for all the targets.

Both monitors types use the Global and Monitor-Specific Configuration (.cfg) Files to configure the required monitor settings. In addition, Multiple-View monitors use the Client Configuration (.clcfg) File. The Client Configuration File enables you to configure different event details for multiple targets.

Event Tracking Methods

EMS Hardware Monitors use one or both of the following methods to track events:

Polling
In polling, a monitor checks the status of its hardware resources at regular intervals, typically every 60 minutes. The monitor generates an event if it detects any unusual behavior. Select the polling interval to provide reasonable detection without impacting system performance. The main disadvantage of polling is that an event is not detected before the next scheduled polling, which renders the system vulnerable to hardware failure.
Asynchronous event detection
Asynchronous detection enables a monitor to detect an event when it occurs, usually during an I/O to the device. Asynchronous event monitoring provides immediate notification and response to a critical situation.

Peripheral Status Monitor

Peripheral Status Monitor (PSM) converts events detected by an EMS Hardware Monitor into changes in hardware resource status. This conversion is required when you use the HP Serviceguard to control package fail over. As the EMS Hardware Monitors detect and report the occurrence of events rather than resource status, a method is required to alert HP Serviceguard when a hardware resource has a status that can impact data availability. The PSM provides this functionality, serving as the interface between the monitors and HP Serviceguard. When an event is generated, the PSM determines whether the event is serious enough to warrant a change in hardware resource status to DOWN. If it is, the PSM alerts the EMS, and the EMS notifies HP Serviceguard about the change in status.

The EMS manages all PSM monitoring requests.

Some monitors can determine when a problem is corrected. These monitors automatically alert the PSM when the hardware they are monitoring is fixed, and the PSM returns the state of the hardware to UP. Other monitors cannot determine when the hardware problem is fixed. Therefore, you must use the set_fixed utility to manually return the operational state to UP. For more information about PSM, see “Peripheral Status Monitor”.

Architecture

Figure 1-1 shows how the EMS components interact.

Figure 1-1 Hardware Monitoring Components

The hardware monitoring process shown in Figure 1-1 works in the following way:

The EMS Hardware Monitor detects an error or an abnormal behavior on the resource it monitors.
The monitor generates an event, which includes suggested corrective action, and communicates it to the EMS framework.
The EMS framework reports the event to the user using the specified notification method.
If the PSM is configured, it processes the events. The PSM changes the device status to DOWN if the event is SERIOUS. PSM notifies the change in device status to the EMS framework, which alerts HP Serviceguard, if configured. The DOWN state causes HP Serviceguard to fail over any package associated with the failed hardware resource.
The system administrator (or the HP service provider) corrects the problem and returns the hardware to its normal operating condition.




	NOTE: Starting with the HP-UX 11i v2 March 2008 release and the HP-UX 11i v3 March 2008 release, SFM is the default monitoring mode. The SFMIndicationProvider replaces some of the EMS Hardware Monitors. For information on the list of monitors that the SFMIndicationProvider replaces, see the SFM Release Notes available at: http://docs.hp.com/en/diag For information on how to switch to the OnlineDiag monitoring mode, see the SFM Administrator's and User's Guide available at: http://docs.hp.com/en/diag

Difference Between Hardware Event Monitoring and Hardware Status Monitoring

The EMS Hardware Monitors detect errors, but they do not retain the details of events. However, the error can persist until it is rectified. For example, a failed disk remains in the same state until it is replaced.

Hardware status monitoring is an extension of event monitoring that converts an event to a change in device status. This conversion, performed by the PSM, provides a mechanism for retaining the event details by storing the resultant status in a log file. This ability provides compatibility with applications such as HP Serviceguard, which require a change in device status to manage high-availability packages.

Products Supported by EMS Hardware Monitors

EMS Hardware Monitors support a variety of system hardware resources. At the time of publication, monitors support the following hardware types:

HP disk arrays, including AutoRAID disk arrays and high-availability disk arrays
HP Fibre Channel Small Computer System Interface (SCSI) Multiplexer
HP Fibre Channel Adapter cards
High-availability storage systems
HP Fibre Channel Arbitrated Loop Hubs
HP Fibre Channel Switch
System memory
Core hardware
Low Priority Machine Checks (LPMCs)
HP-UX kernel resources
HP Fibre Channel Disk Array FC60
SCSI1, SCSI2, and SCSI3 interface cards
System information
HP UPS (Uninterruptible Power Systems
Devices supported by HP device management software (Remote Monitor), for example, virtual arrays

For more information about the supported products, see “Software and Hardware Requirements”.