| Active alarm table
The device shall maintain this table, to maintain all outstanding
(not cleared) alarms generated by the device, including:
1. Not hit device local alarm filter. That means had been reported to
trap receiver (NMS or EMS)
2. Hit local alarm filter. That means not trap is reported to trap receiver
for the alarm.
The device shall maintain two numbers:
1. Alarm sequence number: As described in gponTrapSequenceNumber, it is initially
set to 1 when device restarts, increases per trap reported to trap receiver,
and rollover to 1 once it reaches the maximum value.
From NMS's view, traps received from a nodes will be consecutive, by identifying
the alarm sequence number carried in alarm trap. If breaks, means some trap
may be lost, since packet loss between NMS and NE.
2. Alarm raising number: This is maintained by NE internally and does not intend for
being acknowledged by external manage system. It is designed to uniquely indentify
each alarm raised by NE, excluding duplicates. Once one alarm raised, despite
whether it is hit by alarm filter, NE shall assign a raising number to it. So, it
major functionality is to uniquely identify alarms hit NE local alarm filter, since
this kindly alarm will not be reported to NMS, and hence no alarm sequence number
is assigned to it.
It is initially set to 1 when device restarts, increases per alarm generation (
excluding duplicates), and rollover to 1 once it reaches the maximum value.
FUNCTION DESCRIPTION OF ACTIVE ALARM TABLE:
ALARM RAISING
----------------------
1. When NE detects one fault and associates to one kind of alarm, it shall check whether
there is same alarm in active alarm table, which is indexed by alarm code and alarm
instance. If hits, it means duplicates, update the corresponding entry of the table.
Otherwise,
2. The alarm is not a duplicate. Assign an alarm raising number. Check whether the
alarm hits local alarm filters. If hits, set alarm sequence number to 0.
Create an entry into active alarm table. Otherwise,
3. Assign an alarm sequence number to the alarm. Create an entry into active alarm
table. Report the alarm trap to all trap receivers.
ALARM CLEARING
----------------------
1. When NE detects one fault is cleared and associates to one kind of alarm, it shall
check whether there is corresponding alarm in active alarm table, which is indexed
by alarm code and alarm instance. If no hit, escape this clear event. Upon a hit,
if the alarm sequence number is 0, move the active alarm entry to history alarm
table and update the corresponding fields. At this case, historyAlarmCorrelationId
and historyAlarmSeqNum will be set to 0.
Otherwise,
2. Assign an alarm sequence number to the clear trap. Report the alarm clear trap to
all trap receivers. Move the active alarm entry to history alarm table and update
the corresponding fields. At this case, activeAlarmSeqNum shall be copied to
historyAlarmCorrelationId, and historyAlarmSeqNum shall be set to the new alarm
sequence number for the clear trap.
ALARM SYNCHRONIZATION BETWEEN NE AND NMS
---------------------------------------------
1. Once NMS finds the alarm sequence number in reported alarm traps is not continuous,
it will walk active or history alarm table of the NE for the lost seuqence number.
|
