Glossary
This glossary provides terms and definitions for IBM Cloud Pak® for AIOps.
The following cross-references are used in this glossary:
- See refers you from a nonpreferred term to the preferred term or from an abbreviation to the spelled-out form.
- See also refers you to a related or contrasting term.
While IBM values the use of inclusive language, terms that are outside of IBM's direct influence are sometimes required for the sake of maintaining user understanding. As other industry leaders join IBM in embracing the use of inclusive language, IBM will continue to update the documentation to reflect those changes.
A
accelerator
A set of assets that enable customers to understand core concepts and implement solutions.
AI model
In a machine learning context, a set of functions and algorithms that are trained and tested on a data set to provide predictions or decisions. See also training definition training definition.
alert
A record of an event indicating a fault in the managed environment.
anomaly
A deviation from the expected baseline behaviors.
application
One or more computer programs or software components that provide a function in direct support of a specific business process or processes.
application mapping
Used to expose the on-premises IBM Cloud Pak for AIOps instance to the public cloud. After the application mapping is created, IBM Cloud Pak for AIOps can be accessed from Slack.
auto-remediation
The process of resolving an issue with minimal or no human intervention.
B
boot node
The OpenShift cluster node that is used for running installation, configuration, node scaling, and cluster updates.
C
capability
The core functions of a service. For example, the XYZ service provides "monitoring" capability.
CD
See continuous delivery.
change risk
The risk associated with implementing any type of change, whether it is to code, configurations, or data. See also risk score.
CI
connector
The means by which a data source is connected to a product or service. See also integration.
continuous delivery (CD)
A software development practice that employs techniques such as continuous testing, continuous integration, and continuous deployment so that new features and fixes are packaged and deployed rapidly and at low risk to test environments and then to customers.
continuous integration (CI)
A software development practice where members of a team integrate their work frequently so that there are multiple integrations each day. Integrations are verified by an automated build to detect integration errors as quickly as possible.
coverage
A qualitative determination of the AI's ability to produce insights from your incoming data flow as a result of it's training. See also training definition.
CRD
custom resource
An instance of a custom resource definition. See also custom resource definition.
custom resource definition (CRD)
A customizable YAML file that defines a logically related group of objects in a cluster. Custom resource definitions enable a custom resource to be used like any native Kubernetes object in the cluster. See also custom resource.
D
data set
Data that is extracted and organized from data streams. Data sets serve as inputs for pipelines to train, retrain, or tune models.
data stream
A set of processes that convert extracted data into a normalized data set.
E
edge
The relationship, or link, between resources. See also hop, seed.
error
The fault in an event that causes a failure.
event
A change in state of a service or system.
F
failure
The surfacing of an error to the user.
fault
An error in the internal state of a component in a system.
field mapping
The assignment of values from one integration type to a standard set of values. When one maps source fields, one replaces or supplements those values with new or different values.
H
hop
A step along a single edge from one resource to another. See also edge, seed.
I
incident
An unplanned interruption that causes, may cause, or reduces the quality of an IT service. See also incident.
integration
Any independent application that provides additional capabilities to an IBM service or base product. See also connector.
issue
A cause or potential cause of one or more recurring or similar tickets or incidents.
L
log
Data output from an application or service that describes a transaction or state occurring in that application or service.
M
mean time between failures (MTBF)
The average time between failures.
mean time to failure (MTTF)
The average amount of time before a system lapses into a failed state.
mean time to detect (MTTD)
The average amount of time it takes to identify a failure.
mean time to resolve (MTTR)
The total time period from the start of a failure to when the failure resolves and operations resume.
O
observer
A service that extracts resource information and provides topology data for integrations and localization and blast radius information for events published to ChatOps integrations.
observer job
An automated scheduled task that retrieves topological details for a target system.
operand
The service component used by an operator to perform actions.
operator
A system design that simulates user behavior by linking a controller to custom resources, simulating the act of observing differences in states then acting on them.
P
pipeline
A set of iterative processes that transforms data into insights. For example, retraining pipelines take additional training data sets and improves existing models based on new insights.
proactive channel
A ChatOps channel that is used for change risk notifications. See also reactive channel.
probable cause
A rank-based indicator of causal likelihood that is based on the analysis of alerts and associated topology within the context of an incident.
problem
See issue.
R
reactive channel
A ChatOps channel that is used for outgoing incidents and alerts. See also proactive channel.
resource
A physical or logical component that can be provisioned or reserved for an application or service instance. Examples of resources can include storage, processors, memory, clusters, and VMs.
risk score
A value derived to describe risk across several dimensions associated with change. See also change risk.
RPO (Recovery point objective)
The application loss tolerance, such as the amount of data that can be lost before significant harm to the application occurs.
RTA (Recovery Time Actual)
The actual time that is required for an application to recover completely.
RTO (Recovery Time Objective)
The time that an application can be down without causing significant damage. RTA should be less than RTO.
S
secure tunnel
A feature to package TCP requests and responses, encrypt them with HTTPS, and transport the payloads between IBM Cloud Pak for AIOps and Slack. With Secure Tunnel, you can integrate IBM Cloud Pak for AIOps with Slack without creating firewall rules and policies.
seed
A single resource that serves as the starting point of a topology. Once defined, the topology view is expanded one hop at a time. See also edge, hop.
service
In IBM Cloud Pak solutions, a set of capabilities that provide functionality to a product.
service catalog
A curated repository of services that users can discover and instantiate for use within their organization's applications.
incident
A collection of insights derived from different data sources (logs, events, and alerts) that represent the products' determination of an incident. Incidents build understanding and drive remediation.
T
ticket
A formal record of identified issues or requests that are created against an item and assigned to appropriate users to resolve those issues or complete the requests.
topology
An arrangement of specific, interconnected assets and resources within a network or application.
topology template
A defined topology pattern that is used to dynamically generate topologies based on matching conditions from previously established topologies in the topology database.
training definition
The established model that serves as the baseline for your data to be measured against. For example, your AI model for log anomalies acts as the training definition for all of your log anomaly integrations. See also AI model, coverage.
tunnel worker
Server side of a TCP-over-HTTPS tunnel. It is developed based on OSS project, and is installed in the IBM Cloud Pak for AIOps cluster.
tunnel connection
Used to manage IBM Cloud Pak for AIOps instance that you want to expose. If you want to expose on-premises IBM Cloud Pak for AIOps for Slack integration, first you need to create a tunnel connection.
tunnel connector
Client side of a TCP-over-HTTPS tunnel. It is developed based on OSS project, and can be installed in an OpenShift cluster. Ensure that the OpenShift cluster can be accessed from the public network, and the cluster domain certificate is signed.