Advertisement
4.2 RAS Event Notification
4.2.1 Overview
When an event that must be reported to a user such as hardware failure occurs, this function
notifies the event to an application by setting an event object to the signaled state.
An application can detect an event such as hardware failure by monitoring when the event
object is set to the signaled state.
The event object is reset to the nonsignaled state when the cause of the event is cleared.
4.2.2 Detecting an event
Follow the procedure below to detect an event:
1. Use the OpenEvent Windows API function to get the handle to the event object.
Specify SYNCHRONIZE for the parameter dwDesiredAccess (the access to the event
object).
2. Use the WaitForSingleObject or WaitForMultipleObject Windows API function to
monitor when the event object is set to the signaled state.
Table 4-6 is a list of events to be reported to a user and their respective event objects.
No.
1
A case fan failure occurred.
2
A CPU fan failure occurred.
The temperature inside the chassis became
3
abnormal.
4
SMART anticipated failure in one of the drives.
The power-on (=used) hours exceeded the
5
threshold in one of the drives.
6
The RAID status is normal.
7
The RAID status is not normal.
No. 4: This includes the case that getting the failure prediction status of the drive fails.
Nos. 6 and 7: Apply to D Model only.
No.7: This includes the case that a media error is enabled and notified.
NOTE
When you use an event object in a program, you need to add "Global\" to the
beginning of the name of the event object.
4.2.3 Example of using event objects
We provide a sample program in C (FanErr.c) to show how to monitor event objects. For
information about the name of the sample program and where you can find it, see "6.2
Sample Programs".
Table 4-6 Reported Events
Event
4-11
4. CHECKING THE HARDWARE STATUS
Event object name
HFB_CSFAN_ERR_EVENT
HFB_CPUFAN_ERR_EVENT
HFB_TEMP_ERR_EVENT
HFB_HDD_PREDICT_EVENT
HFB_HDD_OVERRUN_EVENT
HFB_RAID_OPTIMAL_EVENT
HFB_RAID_DEGRADE_EVENT
Advertisement
