Appendix A: Glossary of Terms
Alarm An alarm is a warning about the abnormal conditions of Managed Object (MO).
An alarm always does not represent an error.
Alarm Lifecycle The Alarm Managed Service Object (MSO) associated with an MO depicts the lifecycle of an alarm as:
- An alarm is raised by the alarm service, if the abnormal condition relating to the alarm persists for a time, at least equal to the soaking interval specified. This alarm may be cleared by:
- The alarm service, when it gets a notification from the MO that the abnormal condition which caused the alarm has been cleared.
- RESET, if the MO is hardware, and RESTART, if the MO is software Execution Object (EO).
Alarm Manager The alarm manager enables configuring and handling of alarms. It provides support for alarm soaking, masking, alarm hierarchies, retrieving previous alarm conditions, and correlation of the alarms before publishing.
Alarm Masking Multiple alarms can be raised in a system. Alarm-masking is a procedure that enables the alarm service to publish the alarms with high priority for fault recovery.The current alarm-masking logic implies that all MOs are organized in a fault containment hierarchy which represents the relationship "is physically contained in" for hardware and "runs on" for software. The alarm-masking algorithm masks all alarms of the same or lesser severity level within any sub-tree in the hierarchy.
Alarm Service The OpenClovis alarm service implements a generic engine to process abnormal conditions reported by managed resources. The process includes identifying the alarm type and its severity level and determining whether an alarm has to be raised after running it through soaking and masking procedures.Potential subscribers for alarm events are - OpenClovis Fault Management Service, OpenClovis Availability Management Service and other agents like SNMP, CLI, and so on for reporting alarms to external managers.
Alarm Severity The alarm severity specifies the condition of the service provided by the MO. The severity levels can be critical, major, minor, warning, or cleared and indeterminate (CCITT X.733). The cleared and indeterminate level indicates clearing of one or more previously reported alarms. The critical and major alarms are service-affecting alarms.
Alarm Soaking Soaking is the time defined for extreme conditions before reporting as an alarm. Alarm soaking is possible when a managed resource provides a pair of notifications - one for the occurrence of an extreme condition and another for clearing of the extreme condition. Soaking avoids raising an alarm unless the extreme condition of the alarm persists for a period of time. Sometimes, you may not get a notification for clearing of an extreme condition. In such cases, a poll mechanism can be used to see if the alarm condition is still present after the soaking interval.
Application An Application is customer software built using OpenClovis infrastructure that provides services to the end users. For instance: A SIP server customer application.
SAFplus Platform Console A command-line Interface for debugging allows access to the managed object repository for creating, deleting and otherwise manipulating objects. SAFplus Platform Console also provides viewers for system log and trace data as well as interfaces to individual EOs to view and modify their private (non-persistent) data.
Attributes Attributes are the characteristics or parameters of a Managed Object.
ATCA Chassis An Advanced Telecom Computing Architecture [ATCA] specification is designed to use in the central office grade equipment at the telecommunications sector. The ATCA defines the rack and chassis [shelf] form factors, [passive] backplane, cards, power, and keying. An ATCA card can handle up to 4 PMC daughter cards.
Availability Management Framework (AMF) AMF is a software entity that provides a framework for high availability of applications in a system. It is responsible for instantiating and managing all the OpenClovis SAFplus Platform services. It executes configured recovery actions on the failure of application components. The AMF is built with the close association of two OpenClovis SAFplus Platform components: Component Manager and the Availability Management Service (AMS).
Boot Level OpenClovis SAFplus Platform services and customer applications are mapped to different Boot Levels. The Boot Management Service starts these services (applications) when it reaches the specified BOOT_LEVEL. The BMS can be constrained to boot-up only to a specified BOOT_LEVEL.
BOOT_LEVEL is conceptually similar to the run_level concept in Unix.
Boot Management Service OpenClovis Boot Management Service provides the support for starting or shutting down of the OpenClovis SAFplus Platform services and customer applications on a OpenClovis SAFplus Platform managed platform. The Boot service assumes that an OS has already been booted-up in the target environment. While starting or shutting down a system, the BMS has several BOOT_LEVELS in sequence. At each level, certain services (applications) are started or shutdown.
Boot Profile The boot profile defines a particular type or a particular mode of boot configuration. The list of services performed can be different for different profiles. This helps in obtaining different configurations containing different set of boot levels and different Service Units assigned to them.For Instance, a profile named debug can define a boot configuration, which can be helpful in debugging the boot up process, whereas a profile named production can be used when the desired configuration for normal deployment boot needs to be specified.
Boot State Machine The boot state machine provides a mechanism for a customer to control boot-up (sequences, dependencies, exception handling) using BOOT_LEVELS, RUN-LEVELS, and so on.
Chassis Management Service OpenClovis Chassis Management Service provides support for resource discovery, sensors, and controls on chassis-based hardware platforms. The platforms can be standard-based (ATCA, BladeCenter and so on) or proprietary. The Chassis Management Service can be customized for any platform by providing the platform specific plug-ins.
Checkpoint Service (CPS) CPS provides synchronization of run-time data and context to ensure a seamless failover or switchover of applications. It allows the application to store its internal state and retrieve the information. It also provides a facility for processes to record checkpoint data incrementally and supports non-transparent mode of Checkpointing.
Clovis Object Repository (COR) Clovis Object Repository (COR) is an in-memory object-oriented hierarchical distributed repository of MOs. COR contains the description of each MO and relationships between different MOs. Multiple relationships - hierarchical containment and associations - are supported. COR provides Object lifecycle management, transactions on multiple objects, object change notification, object change propagation and other services.Each OpenClovis SAFplus Platform instance has an instance of the Clovis Object Repository associated with it.The COR instance on the System Director links with COR instances on Blade Directors to provide a single logical system view of all the MOs.
Cold Restart Cold Restart is an element that carries out the entire initialization sequence from the beginning. For example power-on.
COR Persistence The Clovis Object Repository may be persisted using a persistent database. The effect of this is as follows: When a OpenClovis SAFplus Platform instance boots up, COR reads the database and restores the persisted state to each MO. This state is used, for example, by the provisioning service to provision all objects on boot-up.
DBAL Provides a standard interface for any OpenClovis SAFplus Platform infrastructure component or application to interface with the commonly used relational. DBAL currently supports GNU Database Manager.
Default Boot Level During startup, the Component Manager boots the components to a boot level called default boot level as specified in the deployment configuration file. All the components specified up to and including the default boot level are started.
EO Management service The OpenClovis Execution Object (EO) Management Service monitors the health and controls the state of all Execution Objects in the system. State control includes the ability to start, stop, suspend, resume, kill and restart an EO.
EOID Identifier for execution object (EOID is unique within a OpenClovis SAFplus Platform instance).
Error Error is the deviation in the system state or behavior as a result of the use of incorrect data or signal.
Event Events are means by which data may be exchanged between event publishers and event subscribers. An event is characterized by an event channel and event ID.
Event Channel An event channel is a mechanism used by the event service for publishers and subscribers to communicate via events. One or more events (with distinct event ID's) may correspond to an event channel.
Event Filter These are filters used by an EO to specify the events it is interested in.
Event Publisher An EO that publishes an event.
Event Service An OpenClovis SAFplus Platform service that provides a mechanism to publish or subscribe communication based on event channels and asynchronous communication between publishers and subscribers.
Event Subscriber An EO that is interested in receiving published events on a specific event channel.
Execution Object The motivation for the OpenClovis EO is to provide execution contexts independent of process architecture in any OS.Execution objects are programs that implement management interfaces (mandated by OpenClovis SAFplus Platform). These programs allow them to be managed in a OpenClovis SAFplus Platform environment. (Refer to EO Management Service, for details of management functions).EOs may use services provided by other EOs (For example, Checkpointing service). If so, they need to implement client interfaces for the service in question.EOs may provide services to other EOs. In this case they need to implement the service interface.EOs may be made visible to an external manager by representing them as MOs.
Failure A failure in a system occurs when the consumer (human or non-human) of a service is affected by the fact that the system has not delivered the expected service. It is a reflection of unacceptable or incorrect results delivered by a system with respect to a specification. It is an unexpected behavior perceived by the consumer or user of a service.
Fault A fault is a Physical or algorithmic cause of a malfunction. Faults manifest themselves as errors.
Fault Diagnosis Fault Diagnosis is the process of determining the cause of a fault. Fault diagnosis is provided to the granularity of an FRU to support maintainability or serviceability.
Fault Management Service OpenClovis FMS provides a framework for fault management, including fault diagnosis and progressive recovery. Standard recovery methods for software exceptions are available. You can plug-in custom fault recovery methods.
Fault Manager The Fault Manager manages faults in a system and initiates actions. It can handle various user-defined run-time faults, including hardware and software faults. It can prioritize faults to ensure that the critical faults are addressed before the normal or the low-priority faults.The Fault Manager client library notifies alarms to the Fault Manager server located on the same node. The actions to be taken on receiving a fault are controlled by the FM policy associated with the faults.
Field Replaceable Unit (FRU) Field Replaceable Unit is the hardware element of a system that can be replaced by a similar element in the field.
Group Membership Service (GMS) Group membership service provides the facility of leader election. Any application or OpenClovis SAFplus Platform service can register with GMS to keep track of information such as leader change and cluster membership change.
Hardware Abstraction Layer Provides a uniform management interface to hardware peripheral devices via their respective device drivers. Device types include Framer, NP, ASIC, LIU, Switch chip, DSP, and so on. HAL supports the following interfaces: Access, Init, Open, Close, Control, Retrieve, Send, Receive, Download OpenClovis SAFplus Platform. Customer applications are made entirely transparent to configuration changes in the underlying devices by the HAL.
Heartbeat Heartbeat is a message exchanged at regular intervals between two OpenClovis SAFplus Platform instances. A missed heartbeat is the non-arrival of a heartbeat within a timeout period. A configurable number of missed heartbeats are taken to indicate that one of the two instances in question is no longer running or is incommunicable.
Heartbeat Service A service that runs on every OpenClovis SAFplus Platform instance and monitors the health of other OpenClovis SAFplus Platform instances using the heartbeats.
High Availability (HA) High Availability (HA) is used when referring to a system that is capable of providing service most of the time.
Hot Plug Unit An FRU that can be removed or re-inserted even while the system is powered.
Intelligent Object Communication (IOC) OpenClovis Intelligent Object Communication (IOC) provides a transport and OS neutral, fault tolerant communication between OpenClovis Execution Objects using physical links available and user defined transports (E.g.: UDP, TCP, or Ethernet).
Interface Definition Language (IDL) )IDL is a library used by all EOs to communicate efficiently across nodes. Using IDL, OpenClovis SAFplus Platform services can communicate across endian machines and mixed mode (32-bit and 64-bit architecture).
IOC Address Address of a OpenClovis IOC instance that maps to a transport address for each transport or physical link provided.
Link A link is a physical interconnection between nodes. Also referred to as physical medium.
Local Managed Object A managed object abstracting a local resource.
Local Resource A resource that is contained from a fault-containment point of view within a physical node. (Local resources can be software abstractions implemented by programs running on the node, or hardware attached to the node, or the node itself).
Log Service Log service collects, translates, and publishes log messages to record any significant event in the system. For example: operational state change of a component, managed object attribute value change, and so on.
Managed Object (MO) Managed Object provides an abstraction for the manageable properties of a resource in the system. MOs have attributes, support management operations, exhibit behavior and can emit notifications(CCITT X.700).Operations on an MO can be Create or Delete Instances; Get or Modify attributes; Action.Notifications emitted by an MO instance are instance created/deleted; report attribute change; class specific notification such as alarms.Attributes can be single valued, multivalued or grouped in an attribute group (Ex; Chassis attributes relating to blades may be grouped).
Managed Service Object (MSO) MSOs encapsulate the attributes of a Managed Object specific to the particular Object Implement (OI) they are associated with. (Ex: Alarm severity is an attribute related to the alarm MSO; it is a part of the Alarm MSO associated with any MO desiring an alarm service).
Management Information Base (MIB) Management Information Base is a data structure that holds information on how a system is configured or functioning.
Management Interfaces Methods available for managing a system at the boundary of the system and management middleware. In OpenClovis SAFplus Platform, management interfaces (for different management protocols) are provided by the Mediation Library.
Mediation Library The Mediation Library mediates between management agents implementing standard protocols (such as CISCO CLI or SNMP) and the OpenClovis SAFplus Platform to service management requests from respective (CLI, SNMP) management stations. Requests from outside the system are translated to requests on MOs and forwarded to the appropriate MO.
Mean Time to Failure (MTTF) MTTF is the interval in which the system or element can provide service without failure.
Mean Time to Repair (MTTR) MTTR represents the interval in time it takes to resume service whenever a failure occurs.
Middleware OpenClovis SAFplus Platform is often referred to as a middleware. This is to reflect that OpenClovis SAFplus Platform is a layer over the OS, providing system services and APIs to user applications. Applications must be written to OpenClovis SAFplus Platform API's (rather than directly to OS API's) in order to be OS, database and hardware independent and to be manageable and available.
MOID Managed Object Identifier - The MOs have unique global handles associated with them within the system, and using these handles one can address the MOs. The COR handle uses a pair of object class index and object instance index. Object class index uniquely identifies a managed object class and managed object instance index uniquely identifies an instance of a managed object within that class.
MO (Instance) Tree Managed Object Instance Naming Tree - The MO fault containment hierarchy (see OpenClovis Information Model) is used to name instances of objects. For example, in a chassis based system, an instance of a port could be located as follows: Chassis (0)/Blade (3)/Port (2)
MSG-Q A transport supported by the OpenClovis SAFplus Platform infrastructure as an alternative to UDP using message queuing service of the OS. Restricted to use when the sender & the receiver are in the same OS/OpenClovis SAFplus Platform instance.
Name Service The name service facilitates location transparency to the communicating objects by allowing use of object name instead of location specific address. The name service provides name to IOC address translation. A name is user-defined data.
Node Profile Node Profile describes the list of services that can be run on this node and the various attributes of these services. It also describes the characteristics of the node with respect to its role in the chassis. A node can either behave as a Controller Node for the complete chassis or as a normal processing node in the chassis.
Notification Managed Objects emit notifications. Notifications emitted by an MO instance are instance created/ deleted; report attribute-change; class-specific notification such as Alarm (Also referred to as COR notification).
OAMP Refers to Operations, Administration, Maintenance and Provisioning.
OpenClovis SAFplus Platform Services OpenClovis SAFplus Platform services are Execution Objects that provide core-underlying services for OpenClovis SAFplus Platform and applications.
OpenClovis IDE This is a Graphical User Interface (GUI) tool designed to simplify and accelerate the development of System Infrastructure Software for Telecom and other networking products. It enables the customer to rapidly create an information model of the Resources that will be required to manage. It generates customized code for OpenClovis SAFplus Platform, a flagship product of OpenClovis Inc.
OpenClovis Information Model An OpenClovis Information Model (IM) is a generic framework or an abstract representation of the entities in a managed environment.It is a mechanism that describes the characteristics of the network element for which the infrastructure software is being implemented. IM provides a unified view and association of all the physical and logical objects in the managed environment. Information Modeling transports the current content or state of a OpenClovis SAFplus Platform Object such as a blade, port, or any other logical object between different Execution Objects.
Operating System Abstraction Layer (OSAL) The OpenClovis Operating System Abstraction Layer provides commonly used APIs to abstract OS services for OpenClovis EOs and customer applications.The basic OS services include the following:
- Memory Management, Timer Management, Executive Task/Process/thread Management, Signal/Event Management, Resource Management, Messaging services, RTOS task profiling services, synchronization and semaphore management.
- OSAL enables one to easily port the OpenClovis SAFplus Platform and OpenClovis SAFplus-enabled customer applications onto new RTOS environments.
Physical Node A physical node is a particular type of resource that can run a single instance of the operating system and OpenClovis software. (Ex: a blade with a general purpose CPU). Physical nodes may be interconnected via some form of communication medium (Ex: multiple blades via a chassis backplane).The process model describes how OpenClovis SAFplus Platform service EO's and user application EO's map to processes in a specific operating system. (Example: in the multi-process version of Release 1.0 for Linux, all OpenClovis SAFplus Platform EOs run in a single Unix process and user application EOs may be combined in one or more processes).
Pre-provisioning Normally, the provisioning of an object corresponding to an FRU occurs when that FRU is present. However, pre-provisioning takes place when an MO exists and the FRU is absent. Scenarios where this occurs:
- A blade lost power or was extracted from the chassis after being provisioned.
- A user pre-provisioned a blade in the Provision Manager's database in the absence of a blade.
- When the blade becomes active, the provisioning service on the System Controller verifies the provisioned data against the properties of the blade and pushes the data down to the blade if the blade's properties match; otherwise, an identity mismatch alarm will be raised and no data will be pushed down.
Provisioning Service The Provisioning service is an Engine that provisions any type of equipment and facility, virtual or physical in the most generic way. Provisioning is accomplished using the Provisioning EOs and provisioning MSOs.
Reboot Cold reboot occurs when a system is powered off and then powered on. In a warm reboot, the hardware remains powered on and provisioning is not disturbed; the OS is re-started.
Remote Method Dispatch (RMD) Remote Method Dispatch is the foundation for all OpenClovis SAFplus Platform client APIs. It implements remote procedure call semantics using both synchronous and asynchronous methods.
Resource A resource is any entity that can be managed by the OpenClovis SAFplus Platform. A resource is either hardware equipment (for example, a blade) or a software abstraction implemented by programs running on that hardware (for example, a software process).
Routing Table Based on the information in this table, IOC module selects a transport and the next IOC hop, to send a message towards its destination. Multiple routes (to the same destination) are required for fault tolerance.
Rule-Based Engine (RBE) Rule-Based Engine (RBE) provides a mechanism to create rules to be applied to the system instance databased on simple expressions. An expression consists of a mask and a value. These expressions are evaluated on user data and generate a boolean value for the decision.
Run Level Mechanism used to synchronize and order the execution of EOs within and across different OpenClovis SAFplus Platform instances.OpenClovis EOs are configured to execute at some run-level; therefore, they can be controlled by the Boot Manager's Run-Level Controller (RLC). For example, the Alarm Agent is configured at Level 2 after the Provisioning Agent, which is at Level 1. This implies that the Alarm Agent has some execution dependency on the Provisioning Agent and that there is some critical section in the Provisioning Agent's initialization code that needs to execute first before the Alarm Agent can continue its execution.
Shelf It is a structure/frame where one or more chassis are mounted.
Service Service is the output of a system that meets the specification for which the system is devised.
Service ID Uniquely references a service (OpenClovis SAFplus Platform or customer designed).
Simple Network Management Protocol (SNMP) Simple Network Management Protocol (SNMP) is a sub-agent, which provides the flexibility to manage platform and non-platform hardware and OpenClovis SAFplus Platform Information Model. Using SNMP, you can manage the attributes of an MO that includes run a get, set, or notification.
Route Station List List of EOs associated with an COR MSO that need to be visited before an attribute of that MSO is modified.
System The system consists of a collection of physical nodes and programs running on those nodes (Ex: A chassis with multiple blades and running software programs).
System Log A persistent file that records the significant events occuring in the system. Typical entries include events related to boot-up, shutdown, errors, recovery, operator commands, operator log-on and log-off. Each log entry contains a timestamp, reporting entity, affected entity, and severity code and operator identity (if applicable).
Transactions OpenClovis SAFplus Platform supports transactions in the following sense:
- Any service that changes the state of managed objects may provide three methods: validate(), update () and rollback(). All these methods must be successfully executed for the transaction to be complete. The affected MO state is also updated in the persistent COR database as part of the transaction.
- OpenClovis SAFplus Platform also supports complex transactions. A list of component (called route station list) may be associated with each MO/MSO. Each component in the list may provide the three methods listed above. The nested transaction is implemented as follows: The list is traversed top down and the validate method of each successive component is executed. When the last component in the list is reached, the update methods for all the components are executed. If there is any failure happening in the validate() callback, then the rollback () callback is executed for all the components involved in the transaction. The transaction is considered successful if no error is reported by validate method executed as part of transaction. Any error happening in the update phase or the rollback phase is not returned as transaction failure.
Transport OpenClovis term for user (system developer) provided communication facility, which provides IOC access to the physical medium.