Contents

1   Introduction

This document provides Fault Management (FM) information for the Ericsson Centralized User Database (CUDB).

1.1   Document Purpose and Scope

The purpose of this document is to provide a list of application alarms, and describe the alarm management and the application alarm model of the CUDB system. The infrastructure alarms of the system are not in the scope of this document, but are shortly summarized in Section 2.2.

1.2   Revision Information

1.3   Typographic Conventions

Typographic conventions can be found in the following document:

• Typographic Conventions

2   Alarms in the CUDB System

An alarm in the CUDB system is a message sent through the CUDB SNMP interface that informs the operator about a problem in the node which requires attention. The CUDB system can raise two types of alarms:

• Application alarms, raised by CUDB application components, including the software components encapsulated by CUDB.
• Infrastructure alarms, raised independently of the CUDB application.

  Note:  

  For CUDB systems deployed on a cloud infrastructure, then infrastructure alarms are not considered CUDB alarms.

  
  

For more information on the management and alarm model of the application alarms, see Section 2.1. For a brief summary of the management of infrastructure alarms, see Section 2.2.

2.1   Application Alarms

This section describes the management and alarm model of the CUDB application alarms.

CUDB application components (including the software component, that is the operating system and Core Middleware) send their alarms through ESA. The alarms sent to ESA are formatted according to the ERICSSON-SNF-ALARM-MIB and are sent to the Network Management System (NMS). For more information, refer to ESA Fault Management, Reference [90].

2.1.1   Alarm Format and Description

An alarm model is a logical description of the CUDB system described in a tree structure. The alarm model in Figure 1 illustrates the hierarchy of the CUDB application components which are able to raise alarms.

The alarm format used by the CUDB application components is defined by the ERICSSON-SNF-ALARM-MIB. For more information, refer to ESA Fault Management, Reference [90]. The standard location for this file, as well as for other mib files used by the CUDB FM interface, is defined by Ericsson SNMP Agent (ESA) in ESA Setup and Configuration, Reference [91].

Table 1 provides relevant information about the alarms. The Severity, Alarm Event Type and Probable Cause values follow the X.733 International Telecommunications Union (ITU) recommendation, refer to Information Technology - Open Systems Interconnection - Systems Management Alarm Reporting Function ITU-T X.733. CCITT Rec. X.733 (1992 E)Information Technology - Open Systems Interconnection - Systems Management Alarm Reporting Function ITU-T X.733, Reference [92].

2.1.2   Alarm Management

The CUDB application does not provide specific management procedures for the mentioned alarms apart from the manual alarm clearing procedure provided by ESA, as described in Section 2.1.2.1.

2.1.2.1   Clearing Alarms

To clear an alarm, use the fmsendmessage command with the clear parameter:

# fmsendmessage -c <module> <errorcode> <resourceid> [<Alarm Active Description>] <originatingsourceip>

where:

• <module> = Application ID of the alarm.
• <errorcode> = Error Code of the alarm.
• <resourceid> = Active Resource ID of the alarm, prefixed with a dot (".").
• <Alarm Active Description> = Optional parameter that sets the content of the Alarm Active Description. In case it is used, the default Alarm Active Description is replaced with the value of this parameter. For example, it can be used to indicate that the alarm was cleared manually by entering "Manually cleared by User".
• <originatingsourceip> = Originating Source IP of the alarm.

To clear the following non-autocease alarm:

---------------------------------------------------------------
Module                       : STORAGE-ENGINE
Error Code                   : 8
Resource ID                  : .1.3.6.1.4.1.193.169.1.2.8.100
Timestamp First              : Thu Sep 24 13:41:48 CEST 2015
Repeated Counter             : 1
Timestamp Last               : Thu Sep 24 13:41:48 CEST 2015
Alarm Model Description      : Memory usage at Warning level, Storage Engine.
Alarm Active Description     : Storage Engine (DS-group #100): memory usage at Warning level.
ITU Alarm Event Type         : 4
ITU Alarm Probable Cause     : 151
ITU Alarm Perceived Severity : warning
Originating source IP        : 10.143.56.132
Sequence Number              : 554
---------------------------------------------------------------

this command is used:

# fmsendmessage -c STORAGE-ENGINE 8 .1.3.6.1.4.1.193.169.1.2.8.100 "Manually cleared by User" 10.143.56.132

2.1.3   Alarm List

Alarms are grouped by different application components, as shown in Figure 1. They are described in detail in the following subsection. The alarms in the tables are in alphabetical order.

2.1.3.1   Storage Engine

Storage Engine alarms are related to the Database Cluster. The alarm model for PLDB alarms is shown in Figure 2.

The alarm model for DS and general alarms is shown in Figure 3.

Table 2 shows the list of alarms related to Storage Engine.

2.1.3.2   Lightweight Directory Access Protocol Front End

The alarm model for Lightweight Directory Access Protocol (LDAP) Front End (FE)-related alarms is shown in Figure 4.

Table 3 shows the list of alarms related to LDAP FE.

2.1.3.3   Server Platform

The alarm model for Server Platform-related alarms is shown in Figure 5.

Table 4 shows the list of alarms related to Server Platform.

2.1.3.4   Operating System

The alarm model for Operating System-related alarms is shown in Figure 6.

Table 5 shows the list of alarms related to the Operating System.

2.1.3.5   Control

The alarm model for node visibility and global system status related alarms is shown in Figure 7.

Table 6 shows the list of alarms related to Control.

2.1.3.6   Application Counters

The alarm model for Application Counters-related alarms is shown in Figure 8.

Table 7 shows the list of alarms related to Application Counters.

2.1.3.7   Service Availability Forum

The alarm model for Service Availability Forum (SAF)-related alarms is shown in Figure 9.

Table 8 shows the list of alarms related to SAF.

2.1.3.8   Security

The alarm model for Security-related alarms is shown in Figure 10.

Table 9 shows the list of alarms related to Security.

2.1.3.9   Preventive Maintenance

The alarm model for Preventive Maintenance underlaying related alarms is shown in Figure 11.

Table 10 shows the list of alarms related to Preventive Maintenance.

2.1.3.10   Licensing

The alarm model for Licensing - related alarms is shown in Figure 12.

Table 11 shows the list of alarms related to Licensing.

2.1.4   Alarm Relationships

The following alarm relationships are present in the system:

• Storage Engine alarms:
  • If memory usage gets to warning level, the Storage Engine, Memory Usage Too High In PLDB, Warning, Reference [23] alarm is raised. If memory usage then gets to major level, the Storage Engine, Memory Usage Too High In PLDB, Major, Reference [22] is raised, and the previous alarm is cleared. If memory usage continues growing, the Storage Engine, Out Of Memory In PLDB, Reference [27] alarm is raised, and the previous one is maintained.

    The same principle applies for similar alarms related to DS.

  • If the BLOB storage space usage reaches the warning level, the Storage Engine, Tablespace Usage Too High In PLDB, Warning, Reference [44] is raised. If the BLOB storage space usage continues growing until there is no more space left, the alarm Storage Engine, Out Of Tablespace In PLDB, Reference [29] is raised and the previous one is cleared.

    The same principle applies for similar alarms related to DS.

  • If the Consistency Check (refer to CUDB Consistency Check, Reference [88] for more information) finds any inconsistencies between two replicas, the Storage Engine, Data Inconsistency between Replicas Found in PLDB, Minor , Reference [9] or Storage Engine, Data Inconsistency between Replicas Found in PLDB, Major , Reference [8] alarm is raised, depending on the type and amount of inconsistencies. If any of these alarms have been raised earlier before running the Consistency Check, the earlier alarms are cleared, and a new one is raised, if necessary, with the proper severity. These two alarms cannot be raised simultaneously.

    The same principle applies to similar alarms related to the DS.

  • If the Automatic Handling of Network Isolation function is enabled, when a former PLDB master replica rejoins the system as a slave replica (that is, recovery from a system split situation or unexpected mastership change) and it is not able to get in sync with the current master replica, then:
    • The Storage Engine, Unable to Synchronize Cluster in PLDB, Warning, Reference [49] alarm is raised and the Automatic Handling of Network Isolation task is started.
    • After the execution of the first part of the Automatic Handling of Network Isolation task, the Selective Replica Check subtask:
      1. The Storage Engine, Execution of Selective Replica Check Failed, PLDB, Major, Reference [17] alarm is raised if some entries were impossible to retrieve.
      2. The Storage Engine, Automatic Handling of Network Isolation not Completed for PLDB, Reference [2] alarm is raised if Selective Replica Check subtask cannot be completed.
      3. The Storage Engine, Unable to Synchronize Cluster in PLDB, Warning, Reference [49] alarm is cleared. The alarm that follows depends on whether the Self-Ordered Backup and Restore function is enabled, irrespective of whether the Selective Replica Check subtask could be completed and irrespective of its outcome.

         In case the Self-Ordered Backup and Restore function is disabled, the Storage Engine, Unable to Synchronize Cluster in PLDB, Major, Reference [48] alarm is raised, otherwise a new Storage Engine, Unable to Synchronize Cluster in PLDB, Warning, Reference [49] alarm is raised, which indicates the start of the Self-Ordered Backup and Restore process.

    • After the execution of the second part of the Automatic Handling of Network Isolation, the Data Repair subtask:
      1. The Storage Engine, Data Inconsistency between Replicas Repaired, PLDB, Reference [11] alarm is raised if any LDAP entry has been repaired.
      2. The Storage Engine, Unrepaired Data Inconsistency between Replicas, PLDB, Reference [50] alarm is raised if some LDAP entries were impossible to repair.
      3. The Storage Engine, Automatic Handling of Network Isolation not Completed for PLDB, Reference [2] alarm is raised if the Data Repair subtask could not be completed.

      Note:  

      The second part of the Automatic Handling of Network Isolation task, the Data Repair subtask, is started only if the execution of Selective Replica Check subtask was completed.

      
      

    All previously raised alarms, except for the Storage Engine, Unable to Synchronize Cluster in PLDB, Warning, Reference [49] alarm, are maintained.

    The same principles apply to DSs and alarms related to DS.

  • If the Self-Ordered Backup and Restore function is enabled when a slave PLDB replica is unable to get in sync with the current master replica, then:
    • The Storage Engine, Unable to Synchronize Cluster in PLDB, Warning, Reference [49] alarm is raised and the Self-Ordered Backup and Restore task is started.

    In case the Automatic Handling of Network Isolation function is enabled, then raising this alarm is postponed until the alarm Storage Engine, Unable to Synchronize Cluster in PLDB, Warning, Reference [49] related to the execution of Selective Check subtask of Automatic Handling of Network Isolation is cleared.

    • After the execution of the Self-Ordered Backup and Restore task:
      1. The Storage Engine, Unable to Synchronize Cluster in PLDB, Warning, Reference [49] alarm, related to Self-Ordered Backup and Restore task is cleared irrespective of whether the Self-Ordered Backup and Restore task could be completed and irrespective of its outcome.
      2. The Storage Engine, Unable to Synchronize Cluster in PLDB, Major, Reference [48] alarm is raised if the Self-Ordered Backup and Restore task failed to restore the replication.

      Note:  

      Previously raised alarms, except for the Storage Engine, Unable to Synchronize Cluster in PLDB, Warning, Reference [49] alarm, are maintained.

      
      

    Alarms related to the execution of Data Repair subtask of the Automatic Handling of Network Isolation function may be raised in parallel with alarms related to the Self-Ordered Backup and Restore function.

    The same principles apply to DSs and alarms related to DS.

• LDAP FE alarms:
  • The LDAP Front End, Server Down, Reference [56] alarm is linked to each LDAP FE component in the CUDB node. On the other hand, there is a redundancy level indication, corresponding to the maximum number of redundant LDAP FE elements. It is the number of LDAP FEs that can be down without the CUDB node losing its required level of performance.
  • When the number of LDAP FEs down is equal to redundancy level, the alarm LDAP Front End, Processing Redundancy Lost, Reference [55] is raised, while LDAP Front End, Server Down, Reference [56] alarms related to those LDAP FEs are maintained.
  • If one or more LDAP FEs go down, the alarm LDAP Front End, Processing Capacity Below Minimum, Reference [54] is raised, but all previous ones are maintained.

2.2   Infrastructure Alarms

Consider the following regarding infrastructure alarms:

• For CUDB systems deployed on native BSP 8100, the CUDB infrastructure alarms are raised by the underlying BSP 8100 management software. Unlike application ones, these alarms are not sent through ESA agents which are part of the CUDB application. More information on the infrastructure alarms and SNMP configuration can be found in the "BSP Fault Management" and "BSP System Notifications" documents in the BSP 8100 CPI.
• For CUDB systems deployed on a cloud infrastructure, refer to the alarm-related documentation of the infrastructure for more information on the infrastructure alarms and SNMP configuration.

3   Configuration

Some of the alarms are raised when the value of a parameter in CUDB goes above a configured threshold. See the specific alarm OPIs for information on the applicable parameters and thresholds, and how to configure them. It is also possible to configure the NMS IP address where alarm traps are sent. The version used for SNMP is version 3.

For more details about how to configure SNMP for CUDB application components, refer to ESA Fault Management, Reference [91]. For more information on configuring SNMP for infrastructure components, see Section 2.2.

Note:  

In CUDB ESA configuration, the Master Agent Main Port is 60.

Glossary

For the terms, definitions, acronyms and abbreviations used in this document, refer to CUDB Glossary of Terms and Acronyms, Reference [89].

Reference List