1 Introduction
This document describes how to use the Fault Management (FM), Configuration Management (CM), Security Management (SM), File Management (FileM), and Performance Management (PM) in the Common Operation and Maintenance (COM) product.
2 Fault Management
The COM FM allows the operator to detect faults and malfunctions on the system. Based on the information carried in the notifications, the operator can locate the source of the event and the relevant documentation. Based on this information, action can be taken to resolve or prevent failures.
2.1 Basic Concepts
Within the COM FM, the following terms are used: | ||
| Active alarm | An alarm with an alarm state that has been raised but not cleared. | |
| Active Alarm List | The Active Alarm List (AAL) contains all the active alarms in a system using COM and shows the current operational state of the system. The AAL is stored by the persistent storage provided by the Middleware (MW). The operator can, at any time, retrieve all active alarms through the COM Northbound Interface (NBI) using the Command-Line Interface (CLI), Network Configuration (NETCONF), or Simple Network Management Protocol (SNMP). | |
| Alarm | A persistent indication of the fault that has a life cycle or state. An alarm has at least the state raise (initial detection of the fault) or clear (the detection that the fault no longer exists). An alarm can also change state (update) regarding perceived severity. Alarms are also called "stateful alarms" to emphasize that they have a state. | |
| Alert | A discrete alarm without state, an alarm without an associated clear. Alerts are also called "stateless alarms" to emphasize that they have no state. | |
| Heartbeats | The COM FM sends on regular intervals traps to the Management System. A heartbeat trap contains the last sequence numbers and the last event time used for the alarms and alerts. These messages can be used to detect loss of notifications sent as SNMP traps from the COM FM to the Management System. | |
| Notification | A message that can carry an alarm or alert instance. | |
| Trap | An unacknowledged SNMP message that carries a notification or heartbeat. | |
2.1.1 Logs
The COM FM logs all alarm state changes and all alerts using the Log Service Provider Interface (SPI). Each alarm and alert log record is encoded in XML format.
It is the responsibility of the Support Agent (SA) to map the XML log records on physical files.
The XML log record consists of two elements. The first element indicates the time the record is logged. The second element contains the specific information about the alarm or alert and it is formatted as a string separated by semicolons. The format is explained in Table 1.
For a detailed description of each field, refer to the FmAlarm class in Managed Object Model com_fm.
|
Tags and Information |
Description |
|---|---|
|
<FmLogRecord> |
Log record start |
|
<LogTimestamp> |
|
|
Time stamp tag |
Formatted as YYYY-MM-DDThh:mm:ssZ indicates the time the record is logged(1) |
|
</LogTimestamp> |
|
|
<Alarm> or <Alert> |
|
|
Alarm-specific or alert-specific information |
Formatted as a string separated by semicolons containing the following tokens: 1: stateful |
|
</Alarm> or </Alert> |
|
|
</FmLogRecord> |
Log record end |
(1) Z represents the Coordinated
Universal Time (UTC) time zone (equals to +00) or a time zone offset
(+hh:mm or -hh:mm).
Examples of log records in an alarm log are shown in Example 1.
Example 1 Log Records in Alarm Log
<FmLogRecord> <LogTimestamp>2014-03-27T10:53:55+01:00</LogTimestamp> <Alarm>1;2014-03-27T10:53:55+01:00;ManagedElement=1,Equipment=1,⇒ PluginUnit=13;123;429876;HW Fault;418;CRITICAL;Overheated;3;⇒ EQUIPMENTALARM;2014-03-27T10:53:49+01:00;MAJOR;Overheated;1;2;⇒ core1 temp;87;core2temp;93</Alarm> </FmLogRecord> <FmLogRecord> <LogTimestamp>2014-03-27T10:53:55+01:00</LogTimestamp> <Alarm>1;2014-03-27T10:53:55+01:00;ManagedElement=1,Transport=1,⇒ Atm=1;421;792134;Link Overload;613;MAJOR;;5;COMMUNICATIONSALARM;⇒ 2014-03-27T10:53:49+01:00;MINOR;Link Overload;2;1;overload;130</Alarm> </FmLogRecord> <FmLogRecord> <LogTimestamp>2014-03-27T10:53:55+01:00</LogTimestamp> <Alarm>1;2014-03-27T10:53:55+01:00;ManagedElement=1,Transport=1,⇒ Atm=1;421;792134;Link overload;613;CLEARED;;5;COMMUNICATIONSALARM;⇒ 2014-03-27T10:53:49+01:00;MINOR;Link Overload;2;0</Alarm> </FmLogRecord> |
If COM has been configured to use a network Managed Element (ME) identity, this is used in the source Distinguished Names (DNs) in log records as shown in Example 2. For more information about network ME identity, see Section 3.3 Configuration.
Example 2 Log Record in Alarm Log When Using Managed Element Identity
<FmLogRecord> <LogTimestamp>2014-03-27T10:53:55+01:00</LogTimestamp> <Alarm>1;2014-03-27T10:53:55+01:00;ManagedElement=Stockholm,⇒ Equipment=1,Plug=13;123;429876;HW Fault;418;CRITICAL;Overheated;⇒ 3;EQUIPMENTALARM;2014-03-27T10:53:49+01:00;MAJOR;Overheated;1;2;⇒ core1 temp;87;core2 temp;93</Alarm> </FmLogRecord> |
2.2 Configuration
The COM FM can be configured through the COM NBI using the CLI or NETCONF. The use cases are as follows:
- See Section 2.2.1 Configure SNMP Master Agent
- See Section 2.2.2 Configure SNMP Master Agent for DTLS
- Refer to Create SNMPv1 Target
- Refer to Create SNMPv2C Target
- Refer to Create SNMPv3 Target
- Refer to Enable SNMP Target
- Refer to Disable SNMP Target
- Refer to Delete SNMP Target
- Refer to Create SNMP View
- Refer to Check Alarm Status
- Refer to Change Alarm Type Severity
- Refer to Change Heartbeat Interval and see Section 2.5.1 Alarm Information in Multiple Interfaces
For a detailed description of the parameters, refer to Managed Object Model com_fm and Managed Object Model com_snmp.
2.2.1 Configure SNMP Master Agent
The SNMP Master Agent is configured by changing attributes agentAddress in the Managed Object Model com_snmp. A CLI example to set the host and port attributes is shown in Example 3.
Example 3 Configuring SNMP Master Agent
>configure (config)>ManagedElement=1,SystemFunctions=1,SysM=1,Snmp=1,agentAddress (config-Snmp=1-agentAddress)>host=127.0.0.1 (config-Snmp=1-agentAddress)>port=161 (config-Snmp=1-agentAddress)>up (config-Snmp=1)>commit |
2.2.2 Configure SNMP Master Agent for DTLS
The SNMP Master Agent is configured for Datagram Transport Layer Security (DTLS) by changing attributes agentAddressDtls, nodeCredential, and trustCategory in the Managed Object Model com_snmp. Attributes nodeCredential and trustCategory give information about certificates. If those attributes are not configured or invalid, SNMP uses the configuration file to get information about the certificates.
A CLI example to configure the Master Agent for DTLS is shown in Example 4.
Example 4 Configuring SNMP Master Agent for DTLS
>configure (config)>ManagedElement=1,SystemFunctions=1,SysM=1,Snmp=1,agentAddressDtls (config-Snmp=1-agentAddressDtls)>host=0.0.0.0 (config-Snmp=1-agentAddressDtls)>port=10161 (config-Snmp=1-agentAddressDtls)>up (config-Snmp=1)>nodeCredential=⇒ ManagedElement=1,SystemFunctions=1,SecM=1,CertM=1,NodeCredential=nc1 (config-Snmp=1)>trustCategory=⇒ ManagedElement=1,SystemFunctions=1,SecM=1,CertM=1,TrustCategory=tc1 (config-Snmp=1)>commit |
2.2.3 Create SNMPv3 Target for DTLS
To create a SNMPv3 target for DTLS, attributes user, address, and port must be set.
Attribute user must match the rfc822-name, Domain Name System (DNS) address, or IP address in the subjectAltName portion of the certificate presented by the connecting client. The CommonName of the certificate is not used when considering a match, as this is deprecated in the SNMP-TLS-TM Management Information Base (MIB).
A CLI example to create an SNMPv3 target for DTLS is shown in Example 5.
Example 5 Creating SNMPv3 DTLS Target for DTLS
>configure (config)>ManagedElement=1,SystemFunctions=1,SysM=1,Snmp=1,SnmpTargetV3Dtls=1 (config-SnmpTargetV3Dtls=1)>user="user@host.com" (config-SnmpTargetV3Dtls=1)>address="127.0.0.1" (config-SnmpTargetV3Dtls=1)>port=10162 (config-SnmpTargetV3Dtls=1)>up (config-Snmp=1)>commit |
2.3 Act on Received Notification
Each notification received at the Management System has a corresponding alarm Operating Instruction (OPI). The alarm OPI contains information about the notification and the steps needed to eliminate or prevent failures.
The prerequisites for this procedure are as follows:
- All alarm OPIs must be available.
- The ERICSSON-ALARM-MIB document must be available.
To act on a received notification:
- Find the alarm OPI, whose title is equal to the value of either attribute eriAlarmActiveSpecificProblem or eriAlarmAlertSpecificProblem of the received notification.
- Follow the steps in the alarm OPI.
2.4 Ericsson Alarm MIB
The COM FM is compliant with the Ericsson Alarm Management Information Base (MIB) with the following exemptions:
- The alert table, eriAlarmAlertTable, is always empty.
- eriAlarmActiveTableURL always returns an empty string.
- eriAlarmAlertTableURL always returns an empty string.
For more information about the Ericsson Alarm MIB, refer to ERICSSON-ALARM-MIB.
2.4.1 AlarmListRebuilt
To comply with the Ericsson Alarm MIB, the COM FM sends an AlarmListRebuilt notification after a COM restart and after deactivation of alarm suppression. It is an indication to the operator to perform an alarm resynchronization procedure.
For more information about the notification, refer to AlarmListRebuilt.
2.4.2 Alarm Filters
Within the telecom domain, a common problem is that too many alarms are generated. To minimize the number of notifications sent, the COM FM offers two types of alarm filters: alarm toggling filter and transient alarm filter. These filters are by default turned off but can be configured and activated.
2.4.2.1 Alarm Toggling
An alarm is defined as toggling when the same (identical) alarm has been raised and cleared, multiple times, within a defined time.
When the toggling alarm filter is used and an alarm is toggling, alarm attribute additionalText is appended with the text " The alarm is toggling". The appended text is kept until the alarm is cleared.
2.4.2.2 Transient Alarm
Short-term and low-priority alarms are considered transient. A high number of transient alarms cause alarm pollution. If the transient alarm filter is used, alarms considered transient are not raised. For more information about alarm filters, refer to COM Application Developer's Guide.
2.4.3 Alarm Suppression
The main use cases for alarm suppression are planned maintenance tasks where the alarm client risks are flooded with non-important alarms. One example is when a technician is present on-site handling the equipment. All alarms are still available in the AAL, but no alarm notifications are sent.
However, alarms with critical severity, for example, fire alarms, are sent regardless. Active critical alarms that are cleared or assigned a lower severity also generate a notification to indicate that the critical problem is resolved.
For more information on how to turn alarm suppression on/off, refer to COM API and SPI Programmer's Guide.
2.5 Alarm Information Mapping
2.5.1 Alarm Information in Multiple Interfaces
Alarm information is available in the SNMP, CLI, and NETCONF interfaces and in log files, as described in this section.
The heartbeat interval of the COM SNMP alarm interface is configurable as follows:
- SNMP object eriAlarmHbInterval of the ERICSSON-ALARM-MIB
- Attribute heartbeatInterval of the Fm Managed Object (MO) in the COM CLI or NETCONF
A summary of the number of active alarms is available as follows:
- eriAlarmActiveAlarms, eriAlarmSumCritical, eriAlarmSumMajor, eriAlarmSumMinor, and eriAlarmSumWarning SNMP objects of the ERICSSON-ALARM-MIB
- Note:
- The number of alarms of indeterminate severity is available in the SNMP as eriAlarmSumIndeterminate only.
- Attributes totalActive, sumCritical, sumMajor, sumMinor, and sumWarning of MO Fm in the COM CLI and NETCONF
- Note:
- While reading the Fm subtree in a particular transaction, the statistics attributes in Fm (that is totalActive, sumMajor, sumMinor, and so on) cannot always be in sync with the actual FmAlarm instances.
Information about the last issued alarm is available as follows:
- SNMP objects eriAlarmActiveLastSequenceNo and eriAlarmActiveLastChanged of the ERICSSON-ALARM-MIB
- Attributes lastChanged and lastSequenceNo of MO Fm in the COM CLI and NETCONF
Alarms instances are available in multiple interfaces, as summarized in Table 2.
|
Alarm and Alert Log |
FmAlarm MO |
SNMP Object |
SNMP Notification |
|---|---|---|---|
|
eventTime |
lastEventTime |
eriAlarmActiveEventTime, |
eriAlarmActiveEventTime |
|
source |
source |
eriAlarmActiveManagedObject |
eriAlarmActiveManagedObject |
|
majorType |
majorType |
eriAlarmActiveMajorType |
eriAlarmActiveMajorType |
|
minorType |
minorType |
eriAlarmActiveMinorType |
eriAlarmActiveMinorType |
|
specificProblem |
specificProblem |
eriAlarmActiveSpecificProblem |
eriAlarmActiveSpecificProblem |
|
probableCause |
probableCause |
eriAlarmActiveProbableCause |
eriAlarmActiveProbableCause |
|
severity |
activeSeverity |
eriAlarmActiveSeverity, |
The notification type indicates severity information.(1) |
|
additionalText |
additionalText |
eriAlarmActiveAdditionalText, |
eriAlarmNObjAdditionalText, |
|
sequenceNumber |
sequenceNumber |
eriAlarmActiveLastSequenceNo |
eriAlarmActiveLastSequenceNo |
|
eventType |
eventType |
eriAlarmActiveEventType |
eriAlarmActiveEventType |
|
Not applicable |
additionalInfo |
Not applicable |
Not applicable |
|
Not applicable |
fmAlarmId |
Not applicable |
Not applicable |
|
Not applicable |
Not applicable |
eriAlarmActiveResourceId |
eriAlarmNObjResourceId |
(1) The notification types are eriAlarmCritical, eriAlarmMajor, eriAlarmMinor, eriAlarmWarning, eriAlarmIndeterminate, and eriAlarmCleared.
2.5.2 SNMP Objects
Properties of the supported SNMP objects are summarized in Table 3. The deviations are indicated.
|
Object Name and Properties |
Description |
|---|---|
|
ericssonAlarmMIB |
|
|
eriAlarmObjects |
|
|
eriAlarmSummary |
|
|
eriAlarmSumIndeterminate |
This object shows the number of currently active alarms with perceived severity indeterminate. Only stateful alarms are included. |
|
eriAlarmSumCritical |
This object shows the number of currently active alarms with perceived severity critical. Only stateful alarms are included. |
|
eriAlarmSumMajor |
This object shows the number of currently active alarms with perceived severity major. Only stateful alarms are included. |
|
eriAlarmSumMinor |
This object shows the number of currently active alarms with perceived severity minor. Only stateful alarms are included. |
|
eriAlarmSumWarning |
This object shows the number of currently active alarms with perceived severity warning. Only stateful alarms are included. |
|
eriAlarmActiveAlarms |
|
|
eriAlarmActiveNumber |
This object shows the total number of currently active alarms, that is, the total number of entries in the AAL. |
|
eriAlarmActiveLastChanged |
A time stamp when the active AAL was last changed. The value can be used by an operator to initiate an alarm resynchronization procedure. All fields of DateAndTime must be filled out, including the hours and minutes from Coordinated Universal Time (UTC). As such, the value must be 11 octets long. Reference: DateAndTime is defined in RFC 2579. |
|
eriAlarmActiveLastSequenceNo |
The last used sequence number for an alarm state change notification. A Management System can poll this variable to detect lost alarm change notifications. |
|
eriAlarmActiveTableURL |
A URL, in accordance with RFC 4248, pointing to a location where the contents of the alarm table (which is the same as AAL) can be retrieved as a file. The actual file format is out of the scope of this MIB and is found in the system documentation. A string of length zero means that the agent has no URL to provide. Deviation: This object always contains an empty string. |
|
eriAlarmActiveMajorType |
In combination with eriAlarmActiveMinorType, this provides a unique identification of the fault type. Different MO types and instances can share alarm types, but if the same MO reports the same alarm type, it is to be considered as the same alarm state. The alarm type is a simplification of the different X.733 and 3GPP® alarm Integration Reference Point (IRP) alarm correlation mechanisms based on EventType, ProbableCause, SpecificProblem. and NotificationId. In systems where eriAlarmActiveMajorType is not needed for identification purposes, it is not used and must be zero. |
|
eriAlarmActiveMinorType |
In combination with eriAlarmActiveMajorType, this provides a unique identification of the fault type, not including the MO. Different MO types and instances can share alarm types, but if the same MO reports the same alarm type, it is to be considered as the same alarm state. The alarm type is a simplification of the different X.733 and 3GPP alarm IRP alarm correlation mechanisms based on EventType, ProbableCause, SpecificProblem, and NotificationId. |
|
eriAlarmActiveSpecificProblem |
A clear-text unique identification of the alarm type. The mapping between eriAlarmActiveSpecificProblem and [eriAlarmActiveMajorType,eriAlarmActiveMinorType] is one-to-one. |
|
eriAlarmActiveManagedObject |
The 3GPP naming convention must be used as format for the MO parameter. The granularity must be good enough to guarantee unique alarm states and relevant resource identification to the operator. The DN must be "relative" to the nodes "own" root. Reference: 3GPP TS 32.106-8 V3.2, Name convention for Managed Objects. |
|
eriAlarmActiveEventType |
The event type as defined in X.733/X.736. Reference: ITU Recommendation X.733, "Information Technology – Open Systems Interconnection – System Management: Alarm Reporting Function", 1992. |
|
eriAlarmActiveEventTime |
A time stamp of the alarm state change event. This variable represents the last change of the alarm state, such as changed severity or additional text. If the alarm has not changed state, this variable represents the alarm raise time and is the same as originalEventTime. All fields of DateAndTime must be filled out, including the hours and minutes from UTC. As such, the value must be 11 octets long. References: DateAndTime is defined in RFC 2579. ITU Recommendation X.733, "Information Technology – Open Systems Interconnection – System Management: Alarm Reporting Function", 1992. |
|
eriAlarmActiveOriginalEventTime |
The time stamp of the original alarm raises a notification. All fields of DateAndTime must be filled out, including the hours and minutes from UTC. As such, the value must be 11 octets long. Reference: DateAndTime is defined in RFC 2579. |
|
eriAlarmActiveProbableCause |
The probable cause for the alarm originally defined by X.733 and subsequent standards. Refer also to the ERICSSON-ALARM-PC-MIB. Because of the history of problems in maintaining a standardized probable cause, the probable cause is not unique. A best effort mapping of the alarm to existing probable causes is used. Reference: ITU Recommendation X.733, "Information Technology – Open Systems Interconnection – System Management: Alarm Reporting Function", 1992. |
|
eriAlarmActiveSeverity |
The severity of the alarm as defined by X.733. This cannot be the original severity, as the alarm has changed severity. Reference: ITU Recommendation X.733, "Information Technology – Open Systems Interconnection – System Management: Alarm Reporting Function", 1992. |
|
eriAlarmActiveOriginalSeverity |
The original severity as reported by the alarm raise notification. |
|
eriAlarmActiveAdditionalText |
A user-friendly text describing the alarm. The text is both static, depending on the alarm type (probable cause), and dynamic depending on the MO instance and other conditions. This string is longer than the corresponding varbind in the notification to manage large strings. Reference: ITU Recommendation X.733, "Information Technology – Open Systems Interconnection – System Management: Alarm Reporting Function", 1992. |
|
eriAlarmActiveOrigAdditionalText |
A user-friendly text describing the alarm. The text is both static, depending on the alarm type (probable cause), and dynamic depending on the MO and other conditions. This is the original text as reported by the first alarm notification, after which it is not altered. Reference: ITU Recommendation X.733, "Information Technology – Open Systems Interconnection – System Management: Alarm Reporting Function", 1992. |
|
eriAlarmActiveResourceId |
If the alarm refers to an object that is instrumented by the SNMP, this variable indicates the corresponding OID for the MO. Deviation: Not supported. The value is null OID (0.0). |
|
eriAlarmAlerts |
|
|
eriAlarmAlertNumber |
The number of rows in the alert table. |
|
eriAlarmAlertLastChanged |
A time stamp when the alert table was last changed. Can be used by an operator to initiate an update of alerts procedure. All fields of DateAndTime must be filled out, including the hours and minutes from UTC. As such, the value must be 11 octets long. Reference: DateAndTime is defined in RFC 2579. |
|
eriAlarmAlertLastSequenceNo |
The last sequence number used in alert notifications. This can be used to detect a lost notification. |
|
eriAlarmAlertTableURL |
A URL, in accordance with RFC 4248, pointing to a location where the contents of alert table can be retrieved as a file. The file format is out of the scope of this MIB and system-dependant. Deviation: This object always contains an empty string. |
|
eriAlarmAlertTable |
Deviation: The alert table is always empty. |
|
eriAlarmHeartBeat |
|
|
eriAlarmHbInterval |
Notification eriAlarmHeartBeatNotif is sent every eriAlarmHbInterval. Operators can subscribe to the notification using the SNMP framework MIBs by using the snmpNotifyName "heartbeat" (SNMP NOTIFICATION MIB, snmpNotifyTable). |
|
eriAlarmNObjAdditionalText |
A scalar variable, only used in notifications. It is size-ranged with a smaller size than the corresponding type used in the active AAL. This is to ensure that all alarm information fits into one User Datagram Protocol (UDP) packet. The additional text can be further appended by using the dedicated append notification. |
|
eriAlarmNObjMoreAdditionalText |
A scalar variable, only used in notifications. It informs the Management System that the additional text is appended by subsequent append additional text notifications. This object exists because of the limitations of the Protocol Data Unit (PDU) size inherent in the SNMP over UDP, which limits the size of the PDU+headers to under 1500 octets. Otherwise, the notification becomes fragmented. |
|
eriAlarmNObjResourceId |
A scalar variable only, used in notifications. It tells the Management System that there is an SNMP-based resource ID (OID) that identifies the alarming resource. This object exists because of the limitations of the PDU size inherent in the SNMP over UDP, which limits the size of the PDU+headers to under 1500 octets. Otherwise, the notification becomes fragmented. Deviation: Not supported. Always equals 0. |
2.5.3 SNMP Notifications
Properties of the supported SNMP notifications of the ERICSSON-ALARM-MIB are summarized in Table 4. The deviations are indicated.
|
Object Name and Properties |
Description |
|---|---|
|
eriAlarmIndeterminate |
This notification is sent when a resource detects a new alarm state with severity indeterminate. The notification is also used to change the severity or additional text of an alarm, or both. The combination of ManagedObject and MajorType/ MinorType is always unique and can be used by Management Systems to correlate alarm, alarm change, and alarm clear. A corresponding row is created in the AAL. The sequence number increases for every notification and can be used to detect lost notifications. A Management System is to be prepared for appending text to additional text, indicated by the eriAlarmNObjMoreAdditionalText varbind, and sent with eriAlarmAppendInfo. Do not confuse this with a change of additional text. A Management System is also to be prepared to receive a resource ID (OID) identifying the alarming resource if the system sending the notification can provide that information. In that case, eriAlarmNObjResourceId is set to true and the resource ID is sent in an eriAlarmAppendInfo notification. |
|
eriAlarmWarning |
This notification is sent when a resource detects a new alarm state with severity warning. The notification is also used to change the severity or additional text of an alarm, or both. The combination of ManagedObject and MajorType/ MinorType is always unique and can be used by Management Systems to correlate alarm, alarm change, and alarm clear. A corresponding row is created in the alarm table (eriAlarmActiveAlarmTable). Alarm table is the same as AAL. The sequence number increases for every notification and can be used to detect lost notifications. A Management System is to be prepared for appending text to additional text, indicated by the eriAlarmNObjMoreAdditionalText varbind, and sent with eriAlarmAppendInfo. Do not confuse this with a change of additional text. A Management System is also to be prepared to receive a resource ID (OID) identifying the alarming resource if the system sending the notification can provide that information. In that case, eriAlarmNObjResourceId is set to true and the resource ID is sent in an eriAlarmAppendInfo notification. |
|
eriAlarmMinor |
This notification is sent when a resource detects a new alarm state with severity minor. The notification is also used to change the severity or additional text of an alarm, or both. The combination of ManagedObject and MajorType/ MinorType is always unique and can be used by Management Systems to correlate alarm, alarm change, and alarm clear. A corresponding row is created in the alarm table (eriAlarmActiveAlarmTable). The sequence number increases for every notification and can be used to detect lost notifications. A Management System is to be prepared for appending text to additional text, indicated by the eriAlarmNObjMoreAdditionalText varbind, and sent with eriAlarmAppendInfo. Do not confuse this with a change of additional text. A Management System is also to be prepared to receive a resource ID (OID) identifying the alarming resource if the system sending the notification can provide that information. In that case, eriAlarmNObjResourceId is set to true and the resource ID is sent in an eriAlarmAppendInfo notification. |
|
eriAlarmMajor |
This notification is sent when a resource detects a new alarm state with severity major. The notification is also used to change the severity or additional text of an alarm, or both. The combination of ManagedObject and MajorType/ MinorType is always unique and can be used by Management Systems to correlate alarm, alarm change, and alarm clear. A corresponding row is created in the alarm table (eriAlarmActiveAlarmTable). The sequence number increases for every notification and can be used to detect lost notifications. A Management System is to be prepared for appending text to additional text, indicated by the eriAlarmNObjMoreAdditionalText varbind, and sent with eriAlarmAppendInfo. Do not confuse this with a change of additional text. A Management System is also to be prepared to receive a resource ID (OID) identifying the alarming resource if the system sending the notification can provide that information. In that case, eriAlarmNObjResourceId is set to true and the resource ID is sent in an eriAlarmAppendInfo notification. |
|
eriAlarmCritical |
This notification is sent when a resource detects a new alarm state with severity critical. The notification is also used to change the severity or additional text of an alarm, or both. The combination of ManagedObject and MajorType/ MinorType is always unique and can be used by Management Systems to correlate alarm, alarm change, and alarm clear. A corresponding row is created in the alarm table (eriAlarmActiveAlarmTable). The sequence number increases for every notification and can be used to detect lost notifications. A Management System is to be prepared for appending text to additional text, indicated by the eriAlarmNObjMoreAdditionalText varbind, and sent with eriAlarmAppendInfo. Do not confuse this with a change of additional text. A Management System is also to be prepared to receive a resource ID (OID) identifying the alarming resource if the system sending the notification can provide that information. In that case, eriAlarmNObjResourceId is set to true and the resource ID is sent in an eriAlarmAppendInfo notification. |
|
eriAlarmCleared |
This notification is sent when a resource detects a cleared alarm state. The combination of ManagedObject and MajorType/ MinorType is always unique and must be used by Management Systems to correlate alarm and alarm clear. The corresponding row in the alarm table is deleted (eriAlarmActiveAlarmTable). The sequence number increases for every notification and can be used to detect lost notifications. A Management System is to be prepared for appending text to additional text, indicated by the eriAlarmNObjMoreAdditionalText varbind, and sent with eriAlarmAppendInfo. Do not confuse this with a change of additional text. A Management System is also to be prepared to receive a resource ID (OID) identifying the alarming resource if the system sending the notification can provide that information. In that case, eriAlarmNObjResourceId is set to true and the resource ID is sent in an eriAlarmAppendInfo notification. |
|
eriAlarmAppendInfo |
This notification is sent to append further information to an existing alarm. It can be additional text or a resource ID (OID), identifying the alarming resource using an OID. If additional text is sent, do not confuse this with an actual change of additional text, which is reported using the eriAlarm severity notification. A zero-length string value for eriAlarmNObjAdditionalText means that no additional text is sent in this notification. A null OID (0.0) value for eriAlarmActiveResourceId means that no resource ID is sent in this notification. |
|
eriAlarmHeartBeatNotif |
A heartbeat notification with interval according to eriAlarmHbInterval. It contains the last sequence numbers used for alarms and alarm events. These varbinds can be used to detect lost notifications. The notification eriAlarmHeartBeatNotif is sent every eriAlarmHbInterval. Operators can subscribe to the notification using the SNMP framework MIBs by using the snmpNotifyName "heartbeat" (SNMP NOTIFICATION MIB, snmpNotifyTable). |
|
eriAlarmAlarmListRebuilt |
This notification is sent when the AAL has reached a stable situation after a restart or after a system-internal audit process. It is an indication to the operator to perform an alarm resynchronization procedure. Deviation: Not supported. eriAlarmActiveTableURL always returns an empty string. |
In the real notifications, OIDs for all varbinds belonging to eriAlarmActiveAlarmEntry or eriAlarmActiveAlertEntry are appended with the corresponding unique index of alarms or alerts in eriAlarmActiveAlarmTable or eriAlarmActiveAlertTable respectively. For example, OID of varbind eriAlarmActiveManagedObject in an alarm notification corresponding to index 4 in eriAlarmActiveAlarmTable ends with a 4, that is .1.3.6.1.4.1.193.183.4.1.3.5.1.5.4.
The same is applicable for other varbinds belonging to eriAlarmActiveAlarmTable or eriAlarmActiveAlertTable.
3 Configuration Management
The COM CM allows the operator to view and change the configuration data in the system.
3.1 Basic Concepts
Within the COM CM, the following term is used: | ||
| MIM File | The structure of the configuration data is specified in terms of Management Information Model (MIM) files. The MIM XML files correspond to the Managed Object Models (MOMs) that are expressed using the UML®. The MIM XML files comply with the Ericsson-internal Document Type Definition (DTD) named mp.dtd. COM uses the MIM files to validate CM requests. | |
The COM Runtime provides a tool to add, delete, and upgrade model files. For a detailed description, refer to COM Application Developer's Guide.
3.1.1 Logs
COM separates logging of important system events from debugging information. Audit information is also generated. To separate the log information from the debugging information, the NBI Agents use the Log SPI and the Trace SPI offered by COM.
For more information and exact location of the log files, refer to the documentation of the MW SA for each specific MW.
3.2 Managed Object Models
The COM CM function delivers the MOMs specified in Table 5. These MOMs constitute a true subset of the Ericsson Common Information Model (ECIM). Other application-specific MOMs are also accessible from the COM NBI, but these are out of scope of this document.
|
Managed Object Model |
Description |
|---|---|
|
Refer to Managed Object Model com_top. |
Defines the top-level structure of the entire MO class hierarchy. |
|
Refer to Managed Object Model com_sysm |
Defines the model that includes MO classes for the system-level functions. |
|
Refer to Managed Object Model com_fm. |
Defines the model that includes MO classes for the alarm model. |
|
Refer to Managed Object Model com_snmp. |
Defines the model that includes MO classes for the SNMP interface. |
|
Defines the model that includes MO classes for the local authorization model. | |
|
Refer to Managed Object Model com_filem. |
Defines the model that includes MO classes for file management. |
3.3 Configuration
COM System Management can be configured through the COM NBI using the CLI or NETCONF. The use cases are as follows:
- See Section 3.3.1 Change 'ManagedElementId'
- See Section 3.6.1 Configure CLI over SSH
- See Section 3.6.2 Configure CLI over TLS
- See Section 3.6.3 Configure NETCONF over SSH
- See Section 3.6.4 Configure NETCONF over TLS
- Refer to Check General System Attributes
- Refer to List Supported XML Model Files
- Refer to Push XML Model File to Remote File Location
For a detailed description of the parameters, refer to Managed Object Model com_sysm.
3.3.1 Change 'ManagedElementId'
The identity of the root MO, ManagedElement, is set at factory to some default value, normally 1. In a live network, the identity must be set to a unique value provided by the operator. The most important reasons for this are to be able to identify the Network Element (NE) in alarms, PM data, logs, and so on, and to be able to connect the NE MIB into the overall network MIB.
The naming attribute (identity) of an MO cannot be changed after creation, therefore COM provides an additional attribute on ManagedElement called networkManagedElementId. This attribute, if set, is used as the ManagedElement identity, replacing attribute managedElementId. If set to an empty string, the managedElementId is used.
The networkManagedElementId is intended to be used during commissioning of an NE, but the attribute can be changed at any time. If it is changed in runtime, all alarms in the AAL have the source MO updated so, but other data or files that have been generated in the system are not updated.
An example of factory setting is shown in Example 6.
Example 6 Factory Setting
* managedElementId=1 * networkManagedElementId="" * Valid MO DN: ManagedElement=1,SystemFunctions=1,Fm=1 |
An example of live network settings is shown in Example 7.
Example 7 Live Network Settings
* managedElementId=1 * networkManagedElementId=Stockholm * Valid MO DN: ManagedElement=Stockholm,SystemFunctions=1,Fm=1 |
To change ManagedElementId:
- Log on to the COM CLI, refer to Ericsson Command-Line Interface.
- Enter Config mode:
>configure
- Change networkManagedElementId to the desired value:
(config)>ManagedElement=1,networkManagedElementId=<new value>
- Commit changes and return to Exec mode:
(config)>commit
3.4 Configuration Management Using NETCONF
NETCONF is an XML-based protocol that defines operations for accessing and updating a configuration data store.
For detailed information on how to use the NETCONF interface, refer to Ericsson NETCONF Interface.
3.5 Configuration Management Using CLI
The COM CLI is an interface to the Operations, Administration, and Maintenance (OAM) model. The model has a tree-like structure with a ManagedElement as the root, and the subparts and their attributes as child nodes.
When logging on to the CLI from the Secure Shell (SSH), a session is started with the current location placed at a top node, named the ghost node, which is just above the root ManagedElement node.
Nodes are identified by their type name and instance ID. The instance ID is usually (by convention) a number, but can be any string. For example, a field or node storing the serial number of a board can be called serialNumber=1.
A field is equivalent to an attribute. A node is equivalent to an MO.
For detailed information on how to use the CLI, refer to Ericsson Command-Line Interface.
3.6 Management of CM Interfaces
This section describes how to configure CLI and NETCONF services.
3.6.1 Configure CLI over SSH
- Note:
- INSTRUCTION FOR DOCUMENT REUSE: Use this section only if the SSH management feature is supported by the product.
CLI over SSH can be enabled or disabled by changing attribute administrativeState (LOCKED/ UNLOCKED) of the CliSsh class in the Managed Object Model com_sysm.
The behavior of CLI over SSH based on the state of attribute administrativeState is defined as follows:
- If the administrativeState=LOCKED:
- If administrativeState=UNLOCKED:
- Users are allowed to make new connections.
- administrativeState defaults to UNLOCKED.
3.6.2 Configure CLI over TLS
- Note:
- INSTRUCTION FOR DOCUMENT REUSE: Use this section only if the Transport Layer Security Daemon (TLSD) management feature is supported by the product. For configuration of TLSD, refer to Section COM TLSD Manager Configuration File in COM Application Developer's Guide.
COM CLI over TLS is configured for certificates by changing attributes nodeCredential and trustCategory of the CliTls class in the Managed Object Model com_sysm.
The configuration of CliTls is invalid in the following case:
- When attribute trustCategory or nodeCredential is empty and the other refers to respective MO of the CertM MOM.
Example 8 shows how to configure MO CliTls using the CLI.
Example 8 Configuring CliTls MO Using CLI
>configure (config)>ManagedElement=1,SystemFunctions=1,SysM=1,CliTls=1 (config-CliTls=1)>nodeCredential="ManagedElement=1,SystemFunctions=1,SecM=1,⇒ CertM=1,NodeCredential=1" (config-CliTls=1)>trustCategory="ManagedElement=1,SystemFunctions=1,SecM=1,⇒ CertM=1,TrustCategory=1” (config-CliTls=1)>commit |
Example 9 shows how to configure MO CliTls using NETCONF.
Example 9 Configuring CliTls MO Using NETCONF
<?xml version=\"1.0\" encoding=\"UTF-8\"?>
<rpc message-id=\"1\" xmlns=\"urn:ietf:params:xml:ns:netconf:base:1.0\">
<edit-config>
<target>
<running/>
</target>
<config xmlns:xc=\"urn:ietf:params:xml:ns:netconf:base:1.0\">
<ManagedElement xmlns=\"urn:com:ericsson:ecim:ComTop\">
<managedElementId>1</managedElementId>
<SystemFunctions xmlns=\"urn:com:ericsson:ecim:ComTop\">
<systemFunctionsId>1</systemFunctionsId>
<SysM xmlns=\"urn:com:ericsson:ecim:ComSysM\">
<sysMId>1</sysMId>
<CliTls xmlns=\"urn:com:ericsson:ecim:ComSysM\" xc:operation=\"merge\">
<cliTlsId>1</cliTlsId>
<nodeCredential>"ManagedElement=1,SystemFunctions=1,SecM=1,CertM=1,⇒
NodeCredential=1"</nodeCredential>
<trustCategory>"ManagedElement=1,SystemFunctions=1,SecM=1,CertM=1,⇒
TrustCategory=1”</trustCategory>
</CliTls>
</SysM>
</SystemFunctions>
</ManagedElement>
</config>
</edit-config>
</rpc>
]]>]]>
|
3.6.2.1 Administrative State
Attribute administrativeState of the CliTls class controls the CLI connections over Transport Layer Security (TLS). The behavior of CLI over TLS based on the state of attribute administrativeState is defined as follows:
- If administrativeState=LOCKED:
- If administrativeState=UNLOCKED:
- Users are allowed to make new connections.
- administrativeState defaults to LOCKED.
3.6.3 Configure NETCONF over SSH
- Note:
- INSTRUCTION FOR DOCUMENT REUSE: Use this section only if the SSH management feature is supported by the product.
NETCONF over SSH can be enabled or disabled by changing attribute administrativeState (LOCKED/ UNLOCKED) of the NetconfSsh class in the Managed Object Model com_sysm.
The behavior of NETCONF over SSH based on the state of attribute administrativeState is defined as follows:
- If administrativeState=LOCKED:
- If administrativeState=UNLOCKED:
- Users are allowed to make new connections.
- administrativeState defaults to UNLOCKED.
3.6.4 Configure NETCONF over TLS
COM NETCONF over TLS is configured for certificates by changing attributes nodeCredential and trustCategory of the NetconfTls class in the Managed Object Model com_sysm. If these attributes are not configured, the NETCONF TLS component configuration file is used for certificate information.
The configuration of NETCONF TLS is invalid in the following case:
- When attribute trustCategory or nodeCredential is empty and the other refers to respective MO of the CertM MOM.
- Note:
- Credential User Application Programming Interface (API) is required to use certificates from the CertM MOM for NETCONF over TLS. If not found, the NETCONF TLS component configuration file is used for certificates.
Example 10 shows how to configure MO NetconfTls using the CLI.
Example 10 Configuring NetconfTls MO Using CLI
>configure (config)>ManagedElement=1,SystemFunctions=1,SysM=1,NetconfTls=1 (config-NetconfTls=1)>nodeCredential="ManagedElement=1,SystemFunctions=1,⇒ SecM=1,CertM=1,NodeCredential=1" (config-NetconfTls=1)>trustCategory="ManagedElement=1,SystemFunctions=1,⇒ SecM=1,CertM=1,TrustCategory=1” (config-NetconfTls=1)>commit |
Example 11 shows how to configure MO NetconfTls using NETCONF.
Example 11 Configuring NetconfTls MO Using NETCONF
<?xml version=\"1.0\" encoding=\"UTF-8\"?>
<rpc message-id=\"1\" xmlns=\"urn:ietf:params:xml:ns:netconf:base:1.0\">
<edit-config>
<target>
<running/>
</target>
<config xmlns:xc=\"urn:ietf:params:xml:ns:netconf:base:1.0\">
<ManagedElement xmlns=\"urn:com:ericsson:ecim:ComTop\">
<managedElementId>1</managedElementId>
<SystemFunctions xmlns=\"urn:com:ericsson:ecim:ComTop\">
<systemFunctionsId>1</systemFunctionsId>
<SysM xmlns=\"urn:com:ericsson:ecim:ComSysM\">
<sysMId>1</sysMId>
<NetconfTls xmlns=\"urn:com:ericsson:ecim:ComSysM\" xc:operation=\"merge\">
<netconfTlsId>1</netconfTlsId>
<nodeCredential>"ManagedElement=1,SystemFunctions=1,SecM=1,CertM=1,⇒
NodeCredential=1"</nodeCredential>
<trustCategory>"ManagedElement=1,SystemFunctions=1,SecM=1,CertM=1,⇒
TrustCategory=1”</trustCategory>
</NetconfTls>
</SysM>
</SystemFunctions>
</ManagedElement>
</config>
</edit-config>
</rpc>
]]>]]>
|
3.6.4.1 Administrative State
Attribute administrativeState of the NetconfTls class controls the NETCONF connections over TLS. The behavior of NETCONF over TLS based on the state of attribute administrativeState is defined as follows:
- If administrativeState=LOCKED:
- If administrativeState=UNLOCKED:
- Users are allowed to make new connections.
- administrativeState defaults to LOCKED.
3.7 Model File Access
COM provides access to the MIM files loaded by COM. Data related to the MIM files is available through the NBI, according to the Schema class of the Managed Object Model com_sysm. In runtime, one Schema instance exists for each configured MIM file.
The MIM files themselves can be downloaded with the export action of the Schema instances.
4 Security Management
This section describes the SM capabilities of COM and how to use them.
4.1 Overview
The COM SM consists of authorization support of the MOM for users accessing through the NBI. The MOM authorization is based on a combination of roles and rules. User authentication is performed by the SSH.
For an overview of COM SM, see Figure 1.
The flow used for authentication of a user is shown in Table 6.
|
Authentication |
1. SSH authenticates the O&M user by using a Pluggable Authentication Module (PAM). | |||
|
Authorization |
3. The COM NBI agent indicates the start of a new user session to COM SM. | |||
4. COM retrieves the roles of the O&M user from the central Authorization Information Store (AIS). |
4A. The local store is checked first to see if it is an emergency user (refer to Section Create Emergency Group in COM SW Installation Base). COM uses the local database under all conditions. COM considers the Linux /etc/passwd, /etc/group, and /etc/shadow files as the local AIS. | |||
|
4B. If the Access Management (AM) SPI v1 is provided, COM turns to the MW SA to get the roles. | ||||
4C. Otherwise COM SM directs the search to the AM component. COM AM uses the Access Control Service (ACS) User Access Information (UAI) API to retrieve the roles. | ||||
|
5. COM retrieves the rules for the roles of the O&M user from the local AIS. | ||||
|
6. For each O&M operation, the COM NBI agent asks COM SM if access can be granted for the O&M user on the provided operation. | ||||
4.1.1 Authorization Support of MOM for Users Accessing NBI
The roles for a user are stored in authentication data storage.
4.1.2 OS Level Authentication and Authorization
The classical OS-level authentication and authorization is handled by the OS layer. Examples are file-level permission, binary execute permissions, and storage media permissions.
4.1.3 SM Behavior
This section describes the SM behavior from a user perspective.
4.1.3.1 Administrative States
The administrative states are as follows:
- When the SM service is in locked state, no authorization
is performed. Hence all MO operations, MO actions, and CLI commands
are allowed.
- Note:
- The terms "SM service in locked state" and "SM service in unlocked state" means the local state in the SM component based on the administrativeState for the ECIM authorization MO.
- When the SM service is in unlocked state, authorization is performed. The change of the administrativeState affects existing and new sessions.
4.1.3.2 General SM Behavior
The general SM behavior is characterized by the following:
- At the start of every user NBI Agent session, SM data
for that session (user roles, rules) is read from the authorization
data storage.
The SM data is updated during the session if an O&M user with the appropriate privileges changes any or a combination of the following in the Managed Object Model com_secm_local_authorization:
- Attribute administrativeState
- Class CustomRole
- Class CustomRule
- Class Role
- Class Rule
- When the SM MIB is not initialized, the SM service allows any user to execute any NBI Agent operation.
- Authorization exceptions: If any exceptions occur during authorization rule evaluation, such as rule syntax error, the operation is denied. The exception is logged in the system error log by the SM component.
- Access management failure in retrieving roles: If the SM service is unlocked, all MO operations, MO actions, and CLI commands are denied, as role data is not available.
4.2 Security Management Functions
The conceptual view of the user authentication and authorization data is shown in Figure 2.
4.2.1 Fault Management
The SM sends no notifications.
4.2.2 Audit Logging
Audit logging (that is, who logs on and off from the system) is not performed by the SM. The MW or OS layer authenticates the NBI user.
4.2.3 Audit Trail Logging
Audit trail logging (that is, who has done what and when) is not performed by the SM. Instead, it is performed by the NBI Agents, using the Log SPI.
4.2.4 Capacity and Resource Consumption
The resource consumption depends on the size of the role and rule data in the authorization model.
4.2.5 Response Time
The response time depends on the authenticate data storage failover time.
4.3 Roles and Rules
COM delivers role instances for the following:
- SystemAdministrator
- SystemSecurityAdministrator
- SystemReadOnly
- SystemTroubleshooter
- Self
COM also delivers the rule instances connected to the following roles:
- SystemAdministrator
- SystemSecurityAdministrator
- SystemReadOnly
These rule instances only cover the MOM fragments delivered by COM, thus other components and applications must contribute with their own rules. COM does not deliver any rules for the Self and SystemTroubleshooter roles.
For more details on these roles, see Section 4.3.1 Default Roles and Rules.
COM supports implicit navigation through parent fragments. There is no need to specify read access to parent fragments explicitly. Thus the syntax 'MOC1=abc,MOC2=pqr,...,MOCN=xyz$' for ruleData is deprecated. This is illustrated as follows:
ReadWriteExecute access to SecM:
Rule name: SecurityManagement_1
Permission: ReadWriteExecute
Rule data: ManagedElement,SystemFunctions,SecM,*
|
Access is granted as follows:
- Read access to the ManagedElement MO, but not its attributes. Enables navigation in the CLI.
- Read access to the SystemFunctions MO, but not its attributes. Enables navigation in the CLI.
- RWX access to the SecM MO and all MOs below, including all attributes and actions.
4.3.1 Default Roles and Rules
4.3.1.1 System Administrator
A system administrator is responsible for the administration of all non-security-related attributes and capabilities of an ME, for example, ME features, capabilities, configuration parameters, and monitoring of the ME.
The following rules are delivered by COM to provide the system administrator with proper access to model fragments supported by COM:
- Full access to the ManagedElement MO and its attributes.
- Full access to the SystemFunctions MO and its attributes.
- Full access to the Transport MO and its attributes.
- Full access to the Fm MO and all MOs below, including all attributes and actions.
- Full access to the SysM MO and all MOs below, including all attributes and actions.
4.3.1.2 System Security Administrator
A system security administrator is responsible for the administration of the attributes and capabilities of an ME related to security of the ME, regardless of which applications that execute on the ME, for example, ME administrative, user accounts, and authorizations.
The following rules are delivered by COM to provide the system security administrator with proper access to model fragments supported by COM:
- Read access to the ManagedElement MO, but not its attributes. Enables navigation in the CLI.
- Read access to the SystemFunctions MO, but not its attributes. Enables navigation in the CLI.
- Read access to the Fm MO and all MOs below, including all attributes.
- Full access to the SecM MO and all MOs below, including all attributes and actions.
4.3.1.3 System Read Only
A system read only is responsible for the monitoring of system configuration parameters.
The following rules are delivered by COM to provide the system security administrator with proper access to model fragments supported by COM:
- Read access to the ManagedElement MO and its attributes.
- Read access to the SystemFunctions MO and its attributes.
- Read access to the Transport MO and its attributes.
- Read access to the Fm MO and all MOs below, including all attributes.
- Read access to the SysM MO and all MOs below, including all attributes.
4.3.1.4 Self
COM provides one predefined role Self for every user because of which default authorization is enabled for applications. This role is assigned to every user by default.
COM does not deliver any rules for the role Self. Applications that are interested in "Self authorization" can configure application-specific rules under the Self role.
Rules defined under the Role MO instance Self are automatically assigned to the session of a user.
COM supports parameterized and non-parameterized rules under the Self role. See Example 12 for a parameterized rule.
For local users, COM applies both parameterized and non-parameterized rules available under the Self role. For remote users, only non-parameterized rules under the Self role are applied.
Before applying rules under the Self role to a user, COM uses SEC Access Control Service (ACS) User Access Information (UAI) to find whether that user is "local" or not. So, to use the "Self authorization" feature, SEC ACS UAI must be available for COM.
Example 12 Parameterized Rule
RuleName : testParameterizedRule RuleData : ManagedElement=1,SystemFunctions=1,SecM=1,UserManagement=1,UserAccountM=#u,SshPublicKey Permission : RW |
Here, #u is the parameter. For now, username is the only parameter considered. See Table 7.
|
Parameter |
Parameter Description |
Parameter Notation |
|---|---|---|
|
username(1) |
The username of the session. The username is asserted by the service of authentication. |
#u |
(1) This parameter
can give access to MO resources related to user accounts in authentication
methods. To achieve that, the authentication method declares itself
as "local" to authorize the policy decision point for
parameter substitution.
The simplified regular expression of RuleDataType only works for key attributes.
There must be a case-sensitive exact string match between the session parameters and the naming attribute of the specified MO instance.
- Note:
- No roles other than Self must
contain parameterized rules.
COM restricts NBI users to configure the CustomRole MO instance with the role name Self.
4.4 Configuration
The COM SM can be configured through the COM NBI using the CLI or NETCONF. The use cases are as follows:
- Refer to Unlock Local Authorization Method and see Section 4.4.1 Change Administrative State Using NETCONF
- Refer to Lock Local Authorization Method
- Refer to Create Custom Role
- Refer to Change Custom Role
- Refer to Delete Custom Role
- Refer to Create Custom Rule
- Refer to Change Custom Rule
- Refer to Delete Custom Rule
- Refer to View Roles and Rules
For a detailed description of the parameters, refer to Managed Object Model com_secm_local_authorization.
The prerequisites for these procedures are as follows:
- The user must be assigned a system authentication role having sufficient privileges to access the SM branch.
- ManagedElement , SystemFunctions , SecM , UserManagement, and LocalAuthorizationMethod must exist with instance ID = 1. Attribute networkManagedElementId of ManagedElement must be unset or set to 1.
4.4.1 Change Administrative State Using NETCONF
To change administrativeState using NETCONF, in this case setting the SM authorization to state UNLOCKED:
Example 13 Changing Administrative State Using NETCONF
<?xml version="1.0" encoding="UTF-8"?>
<rpc message-id="1" xmlns="urn:ietf:params:xml:ns:netconf:base:1.0">
<edit-config>
<target>
<running/>
</target>
<config xmlns:xc="urn:ietf:params:xml:ns:netconf:base:1.0">
<ManagedElement xmlns="urn:com:ericsson:ecim:ComTop">
<managedElementId>1</managedElementId>
<SystemFunctions>
<systemFunctionsId>1</systemFunctionsId>
<SecM xmlns="urn:com:ericsson:ecim:ComSecM">
<secMId>1</secMId>
<UserManagement>
<userManagementId>1</userManagementId>
<LocalAuthorizationMethod ⇒
xmlns="urn:com:ericsson:ecim:ComLocalAuthorization" xc:operation="merge">
<localId>1</localId>
<administrativeState>UNLOCKED</administrativeState>
</LocalAuthorizationMethod>
</UserManagement>
</SecM>
</SystemFunctions>
</ManagedElement>
</config>
</edit-config>
</rpc>
|
5 File Management
FileM is a service implementing parts of the ECIM fragment FileM and it is characterized by the following:
- Provides a way for file producers and file consumers on the ME to organize and maintain (housekeep) files and folders in the file system.
- Provides and facilitates services to expose files and
folders to NBI services, for example, the SSH File Transfer Protocol
(SFTP), CLI, and NETCONF.
- Note:
- COM supports DN-restricted characters "+", "*", and "#" in the FileGroup value, which are displayed as \2B, \2A, and \23 in the NBI,
where 2B, 2A, and 23 are the hexadecimal ASCII code of these characters,
except these all the other DN-restricted characters (<, > , ; and
so on) in the FileGroup value are displayed
as " ??" in the NBI. All the
DN-restricted characters (+, < , >, #, and so on) in the Filenames
are displayed as " ??" in the
NBI.
For MO name convention and a full list of DN-restricted characters, refer to 3GPP TS 32.300 V9.0.0.
All access to files and folders through the NBI is authorized by rules defined in the ECIM SecM fragment, see Section 4 Security Management.
5.1 Configuration
The COM FileM can be configured through the COM NBI using the CLI or NETCONF. The use cases are as follows:
- Refer to Fetch File in Logical File System
- Refer to Delete File in Logical File System
- Refer to List File Groups and File Information in Logical File System
- Refer to Configure Preventive Maintenance Policy Deleting Files in Logical File System
- Refer to Configure Preventive Maintenance Policy Reporting Alarms for Logical File System
For a detailed description of the parameters, refer to Managed Object Model com_filem.
6 Performance Management
3GPP TS 32.401 V10.0.0 defines PM as follows:
"Any evaluation of a system’s behavior requires performance data collected and recorded by its NEs according to a schedule established by the NE manager. This aspect of the management environment is termed Performance Management. The purpose of any Performance Management activity is to collect data, which can be used to verify the physical and logical configuration of the network and to locate potential problems as early as possible."
TS 32.401 also defines the different phases in the process of performance measurement administration. COM is responsible of two of them: measurement job administration (section 4.2.1 in the TS) and local storage of the results at the NE (section 4.2.3 in the TS). Generation of the performance measurement results is the responsibility of the MW system.
6.1 Basic Concepts
3GPP TS 32.401 describes all the PM concepts. The most relevant concepts from a COM user point of view are as follows:
| Job granularity period | Time between the initiation of two successive gatherings of measurement data triggered by this job. | |
| Job reporting period | Time between the generations of two report files including measurement data generated by this job. | |
| Measurement job | Process executed in the NE to accumulate measurement result data and assemble it for collection or inspection, or both. | |
| ROP report file | The Result Output Period (ROP) report file is an XML file compliant with the 3GPP TS 32.432 V10.0.0. The file contains the results of one or more measurement collections triggered by one or more measurement jobs along one or more Granularity Periods (GPs). | |
6.2 Measurement Job Administration
The ECIM PM fragment provides support to perform all the tasks related with the administration of measurement jobs. The Pm class is present under SystemFunctions and this is present under ManagedElement.
6.2.1 PmJob
Measurement jobs are represented as instances of PmJob defined in the ECIM PM fragment. Administration of jobs (creation, deletion, and modification of job attributes) is made by accessing the ECIM through the NBI (through the CLI or NETCONF) and instantiating and deleting PmJob MOs, or changing the values of the attributes in these MOs.
COM requires ECIM PM Version 1.2 or higher. For information about the meaning of the attributes and possible values, refer to ECIM PM Documentation.
Deviations in COM from ECIM PM Version 1.2 are as follows:
- COM PM assumes that the Job Reporting Period is equal to the Job GP.
- COM PM does not support job priority.
- COM PM does not limit the number of measurements in a report.
6.2.2 Configure PmGroup
Several instances of PmGroup can be found under the Pm class. They describe the measurements that are available on the NE.
According to the ECIM PM fragment documentation "The PmGroup is a grouping of the measurements into logical grouping". Each PmGroup has one or more MeasurementType instances.
To add a measurement to a certain job:
- Create a MeasurementReader MO under the PmJob MO.
- Set the members of struct measurementSpecification so they contain a reference to the PmGroup and MeasurementType to be added to the job.
6.2.3 PmMeasurementCapabilities
The PmMeasurementCapabilities class of the ECIM PM fragment contains attributes describing the measurement capabilities of the current NE. For details of the meaning of each attribute, refer to ECIM PM Documentation.
6.3 Storage of Performance Measurement Results
Once the measurement jobs are set up, the MW system is responsible of collecting the measurement results for every job. Once the results are ready, at the end of the GP of every measurement job, the COM PM functionality writes results into Result Output Period (ROP) report files.
6.3.1 Location of ROP Files
The location where the ROP files are stored can be retrieved from attribute fileLocation in the PmMeasurementCapabilities class.
If FileM is used:
- fileLocation is the relative path /PerformanceManagementReportFiles.
- fileGroup is the DN of a FileGroup MO.
If FileM is not used:
- fileLocation is the absolute path, for example: /var/filem/internal_root/PerformanceManagementReportFiles.
- fileGroup is empty.
The COM PM functionality performs housekeeping of PM report files after each GP ends. The number of reports published does not exceed the value of attribute maxNoOfPmFiles in PmMeasurementCapabilites.
- Note:
- It is not allowed to set up FileM housekeeping policies of PM report files or the PM report files directories. Such a configuration leads to undefined behavior in the system. Attribute fileGroup is deprecated from ECIM PM 2.0 onwards.
6.3.2 Content of ROP Files
The PM report file format is according to 3GPP TS 32.432 type A and is built on the example in 3GPP TS 32.435 Annex A.2. COM also supports multivalue counters, which are not part of the 3GPP specification.
COM PM supports job grouping. This is indicated by attribute jobGroupingSupport in PmMeasurementCapabilities. The attribute is set to true during COM PM startup. ROP files are written as follows:
- If attribute jobGroup in the PmJob MO is present, a separate report file is generated containing all measurements for each PmJob containing this jobGroup. The <UniqueId> part of the filename is appended with " _<jobGroup>".
- If attribute jobGroup in the PmJob MO is absent, the default behavior is to include the measurement results for all such jobs in a single PM report file per GP, containing the results of those jobs for which no jobGroup tag was assigned.
If a MeasurementType instance (counter) is measured in more than one PM Job, the measurement result is reported separately for each job in a measurement report file.
If COM does not support ECIM PM 2.1, attribute jobGroupingSupport is not set and all measurement results are written to a single PM report file per GP.
6.3.3 Job Reporting Period
COM does not support measurements from several GPs in one ROP file. Attributes reportingPeriod and granularityPeriod of the PmJob MO are used to write the report. COM PM expects these attributes to have the same value. The validation of these attributes is expected to be done at job creation time by the PM service in the SA/MW. COM PM sets attribute fileRPSupported=false in the PmMeasurementCapabilities class during startup. This signals to the SA/MW that COM PM does not support different values to reportingPeriod and granularityPeriod for each individual PmJob when fileRPSupported=false and the SA/MW can validate this
Example
The collection of measurement results is started when the time is 0:00. Six jobs are set up, two (job 1 and 2) have a GP of one minute, two (job 3 and 4) have a GP of five minutes, and two (job 5 and 6) have a GP of 15 minutes. The creation of the files is as shown in Table 8.
|
Time |
Number of Created Files |
Content |
|---|---|---|
|
Between 0:00 and 0:01 |
0 |
No files are created. Measurement time collection has started at 0:00 and the GP has not ended. |
|
Between 0:01 and 0:02 |
1 |
Measurement results for jobs 1 and 2 during period (0–1). |
|
Between 0:02 and 0:03 |
1 |
Measurement results for jobs 1 and 2 during period (1–2). |
|
Between 0:03 and 0:04 |
1 |
Measurement results for jobs 1 and 2 during period (2–3). |
|
Between 0:04 and 0:05 |
1 |
Measurement results for jobs 1 and 2 during period (3–4). |
|
Between 0:05 and 0:06 |
2 |
Measurement results for jobs 1 and 2 during period (4–5). |
|
Measurement results for jobs 3 and 4 during period (0–5). | ||
|
Between 0:06 and 0:07 |
1 |
Measurement results for jobs 1 and 2 during period (5–6). |
|
Between 0:10 and 0:11 |
2 |
Measurement results for jobs 1 and 2 during period (9–10). |
|
Measurement results for jobs 3 and 4 during period (5–10). | ||
|
Between 0:14 and 0:15 |
1 |
Measurement results for jobs 1 and 2 during period (13–14). |
|
Between 0:15 and 0:16 |
3 |
Measurement results for jobs 1 and 2 during period (14–15). |
|
Measurement results for jobs 3 and 4 during period (10–15). | ||
|
Measurement results for jobs 5 and 6 during period (0–15). |
One minute is the upper time bound that it takes to get the results for a job written into a file after the end of the GP. After this time, the absence of an expected report file or the absence of measurement data for a configured active job is to be considered as "suspect". This situation can be because of failures in the PM service in the SA/MW or because of file operation errors, in which case the COM PM functionality logs an error.
6.4 ECIM PM Model Compliance
COM does not deliver the PM model. However, the ownership of the fragment is split between COM and the SA/MW. The ownership of COM lies in the ROP file-related attributes of the PmMeasurementCapabilities class. For more information, refer to COM ECIM Statement of Compliance.

Contents

