There exists a class of systems that should implement dynamic severity assignment. A 'pedestal' is typically a dark-green metal box mounted on a concrete or stone foundation in which carrier-class companies house equipment.
The central office typically controls the temperature and humidity of the environment, reducing reliance on a system's fans. Thus, the customer probably has a desire to reduce the severity of alarms indicating the failure of a fan. However, a pedestal environment has a much greater reliance on a system's fans. Thus, the customer probably has a desire to increase the severity of alarms indicating the failure of a fan.
An alarm filter profile controls which alarm types the agent will monitor and signal for the corresponding physical entity. If the value of this object is '0', then the agent monitors and signals all alarms associated with the corresponding physical entity. A value of '0' indicates that there the corresponding physical entity currently is not asserting any alarms.
Note, an alarm indicates a condition, not an event. An alarm has two states: 'asserted' Indicates that the condition described by the alarm exists. For example, a slot in a chassis may define an alarm that specifies whether the slot contains a module. At the time of module insertion, the physical entity corresponding to the slot asserts this alarm, and the alarm remains asserted until the slot becomes empty. If an alarm is being asserted by the physical entity, then the corresponding bit in the alarm list is set to a one.
Observe that if the physical entity is not currently asserting any alarms, then the list will have a length of zero. The value of this object starts at '1' and monotonically increases for each alarm condition transition monitored by the agent. If the value of this object is '', the agent will reset it to '1' upon monitoring the next alarm condition transition.
A management client can create a conceptual row in this table by setting this object to 'createAndWait' or 'createAndGo'. If a request to create a conceptual row in this table fails, then the system is not capable of supporting any more alarm filters.
Before modifying a conceptual row in this table, the management client must set this object to 'notInService'. After modifying a conceptual row in this table, the management client must set this object to 'active'. This operation causes the modifications made to an alarm filter profile to take effect.
An implementation should not allow a conceptual row in this table to be deleted if one or more physical entities reference it. On the first instantiation of an alarm filter profile, the value of this object is a zero-length string. However, an agent may choose to set the value to a locally unique default value. If an implementation supports write access to this object, then the agent is responsible for ensuring the retention of any value written to this object until a management client deletes it.
The level of retention must span reboots and reinitializations of the network management system, including those that result in different assignments to the value of the entPhysicalIndex associated with the physical entity.
Table ceAlarmDescrMapTable. Observe that this table is sparse in nature, as it is rarely the case that a physical entity type needs to define every alarm in its alarm space. Most network devices and programs ship with so-called MIB files to describe the parameters and meanings i. Bellcore TR-NWT defines these severities as follows: 'critical' An alarm used to indicate a severe, service- affecting condition has occurred and that immediate corrective action is imperative, regardless of the time of day or day of the week.
These troubles require the immediate attention and response of a technician to restore or maintain system capability. The urgency is less than in critical situations because of a lesser immediate or impending effect on service or system performance.
Observe that it is not necessary that all the alarms within a space be defined. The bits in the first octet represent alarm types identified by the integer values 1 through 8, inclusive, The bits in the second octet represent alarm types identified by the integer values 9 through 16, inclusive, and so forth. The least significant bit of an octet represents the alarm type identified by the lowest integer value, and the most significant bit represents the alarm type identified by the highest integer value.
The figure shown below illustrates the format of an alarm list. A special case is an alarm list having a length of '0', which represents an alarm list of all zeros. Observe that this table is sparse in nature, as it is rarely the case that a physical entity type needs to define every alarm in its alarm space. An implementation may chose to not allow dynamic severity assignment, in which case it would restrict access to this object to be read-only. If an implementation allows dynamic severity assignment, then a management client can revert to the default severity by writing the value '0' to this object.
There exists a class of systems that should implement dynamic severity assignment. A 'pedestal' is typically a dark-green metal box mounted on a concrete or stone foundation in which carrier-class companies house equipment.
The central office typically controls the temperature and humidity of the environment, reducing reliance on a system's fans. Thus, the customer probably has a desire to reduce the severity of alarms indicating the failure of a fan. However, a pedestal environment has a much greater reliance on a system's fans. Thus, the customer probably has a desire to increase the severity of alarms indicating the failure of a fan.
Reading this object should always result in a value of 'false'. Observe that alarm cutoff does not have an effect on monitoring, history logging, generation of notifications, or syslog message generation. Not every physical component will have a asset tracking identifier, or even need one.
An agent does not have to provide write access for such entities, and may instead return a zero-length string. If write access is implemented for an instance of entPhysicalAssetID, and a value is written into the instance, the agent must retain the supplied value in the entPhysicalAssetID instance associated with the same physical entity for as long as that entity remains instantiated.
If no asset tracking information is associated with the physical component, then this object will contain a zero- length string. If this object contains the value 'true 1 ' then this entPhysicalEntry identifies a field replaceable unit.
For all entPhysicalEntries which represent components that are permanently contained within a field replaceable unit, the value 'false 2 ' should be returned for this object. For agents which implement more than one naming scope, at least one entry must exist. Agents which instantiate all MIB objects within a single naming scope are not required to implement this table.
Entities may be managed by this agent or other SNMP agents possibly in the same chassis. The value should be a small positive integer; index values for different logical entities are are not necessarily contiguous. This object should contain a string which identifies the manufacturer's name for the logical entity, and should be set to a distinct value for each version of the logical entity.
The agent should allow read access with this community string to an appropriate subset of all managed objects and may also return a community string based on the privileges of the request used to read this object. Note that an agent may return a community string with read-only privileges, even if this object is accessed with a read- write community string.
However, the agent must take care not to return a community string which allows more privileges than the community string used to access this object. A compliant SNMP agent may wish to conserve naming scopes by representing multiple logical entities in a single 'default' naming scope. This is possible when the logical entities represented by the same value of entLogicalCommunity have no object instances in common. For example, 'bridge1' and 'repeater1' may be part of the main naming scope, but at least one additional community string is needed to represent 'bridge2' and 'repeater2'.
Logical entities 'bridge1' and 'repeater1' would be represented by sysOREntries associated with the 'default' naming scope. This object may also contain an empty string if a community string has not yet been assigned by the agent, or no community string with suitable access rights can be returned for a particular SNMP request.
Note that this object is deprecated. SNMPv3 agents may return a zero-length string for this object, or may continue to return a community string e. This object, together with the associated entLogicalContextName object, defines the context associated with a particular logical entity, and allows access to SNMP engines identified by a contextEngineId and contextName pair.
If no value has been configured by the agent, a zero-length string is returned, or the agent may choose not to instantiate this object at all. This object, together with the associated entLogicalContextEngineID object, defines the context associated with a particular logical entity, and allows access to SNMP engines identified by a contextEngineId and contextName pair.
For each logical entity known by this agent, there are zero or more mappings to the physical resources which are used to realize that logical entity. An agent should limit the number and nature of entries in this table such that only meaningful and non-redundant information is returned.
For example, in a system which contains a single power supply, mappings between logical entities and the power supply are not useful and should not be included.
Also, only the most appropriate physical component which is closest to the root of a particular containment tree should be identified in an entLPMapping entry. For example, suppose a bridge is realized on a particular module, and all ports on that module are ports on this bridge. A mapping between the bridge and the module would be useful, but additional mappings between the bridge and each of the ports on that module would be redundant since the entPhysicalContainedIn hierarchy can provide the same information.
If, on the other hand, more than one bridge was utilizing ports on this module, then mappings between each bridge and the ports it used would be appropriate. Also, in the case of a single backplane repeater, a mapping for the backplane to the single repeater entity is not necessary.
Note that the nature of the association is not specifically identified in this entry. It is expected that sufficient information exists in the MIBs used to manage a particular logical entity to infer how physical component information is utilized.
Each physical port in the system may be associated with a mapping to an external identifier, which itself is associated with a particular logical entity's naming scope. A 'wildcard' mechanism is provided to indicate that an identifier is associated with more than one logical entity.
Note that only entPhysicalIndex values which represent physical ports i. If this object has a non-zero value, then it identifies the logical entity named by the same value of entLogicalIndex.
If this object has a value of zero, then the mapping between the physical component and the alias identifier for this entAliasMapping entry is associated with all unspecified logical entities. For example, to indicate that a particular interface e. In this example, all logical entities except 4, 5, and 10, associate physical entity 33 with ifIndex. Since only physical ports are modeled in this table, only entries which represent interfaces or ports are allowed.
If an ifEntry exists on behalf of a particular physical port, then this object should identify the associated 'ifEntry'.
0コメント