This topic lists basic SOFAStack terms to help you better understand the product.
A | B | C | D | F | G | H | I | J | K | L | M | O | P | Q | R | S | T | W | X | Y | Z
A
AKE
Ant Financial Kubernetes Engine (AKE) is a container engine that partitions and abstracts underlying physical resources, such as computing, network, and storage. AKE uses Kubernetes and Docker to pool all physical resources. It provides pay-as-you-go computing, network, and storage resources to higher-level services.
AKE Plus
Ant Financial Kubernetes Engine Plus is a container cloud platform developed by Ant Group based on open source Kubernetes and containerd. It is used to deploy Alibaba Finance Cloud products. It provides unified abstractions for computing, storage, and network resources and masks the differences between environments, such as physical servers and virtual machines.
AKS
Application Kubernetes Service (AKS) fully integrates with Kubernetes to provide platform capabilities such as cluster management, authentication and authorization, container networking, and persistent volume storage. For large-scale distributed systems in finance scenarios, AKS provides a variety of release policies to meet different needs. It helps you smoothly transition from traditional architectures, meet financial technology risk assurance requirements, and implement container services in large-scale, finance-grade operations and maintenance (O&M) scenarios.
ALB
Ant Financial Load Balancer (ALB) is a traffic distribution control service that distributes access traffic to multiple backend servers based on forwarding rules. Traffic distribution expands the external service capabilities of application systems and improves their availability by eliminating single points of failure.
API
An application programming interface (API) is a set of predefined functions or a convention for connecting different components of a software system.
API groups are logical groupings of APIs that share the same group identity for fencing.
B
Label
A label is a native Kubernetes concept used to tag resources for aggregation or matching.
Secret
A secret is a native Kubernetes concept used to store encrypted user content.
Report
A report is a view that contains one or more data sources. It centrally displays the monitoring results for each data source.
Sidecar
A sidecar is a lightweight network agent deployed with an application. It intercepts all inbound and outbound network requests to implement various network policies, such as service discovery, load balancing, traffic splitting, fault injection, circuit breaking, and phased releases.
Message type
The media type information in requests and responses. It tells the server how to process request data and tells the client, usually a browser, how to parse response data.
Message encoding
The character set of the text content that the client accepts.
Deployment Unit
A cell is a self-contained collection that can perform all business operations. It contains all the services required for the business and the data allocated to that cell. A unit-based architecture uses cells as the basic deployment unit. Multiple cells are deployed across all data centers on the site. The number of cells in each data center is variable. Each cell deploys all applications required by the system. The data is a subset of the full data, partitioned by a specific dimension.
C
Cron expression
A Cron expression is a string with a fixed format that specifies a point in time or a recurring interval for a trigger. For more information, see Cron expression details.
Consumer instance
A consumer instance is an object instance of a consumer. Different consumer instances can run in different processes or on different machines. A thread pool is configured within a consumer instance to consume messages.
Storage Class
A storage class is a native Kubernetes concept. It is usually defined by a system administrator to specify supported storage categories. Different categories can have different Service-level Agreements (SLAs), backup policies, and other characteristics.
Persistent volume claim (PVC)
A persistent volume claim (PVC) is a native Kubernetes concept. A PVC is bound to a persistent volume (PV).
Consumer offset resetting
Consumer offset resetting resets the consumption progress of a consumer for a subscribed topic based on a timeline. This is done within the message persistence period, which is three days by default. After the reset, the consumer receives messages sent by the producer to the message queue server after the specified point in time. For more information, see Reset consumer offsets.
Error rate
The proportion of errors relative to the number of requests for an application within a specified period.
Participant
A participant provides branch transaction services. When a participant is invoked by an initiator, it becomes part of the distributed transaction started by that initiator and acts as a branch transaction. A distributed transaction can have multiple participants.
Repository
A repository specifically refers to a code repository.
D
DataId
A string that globally and uniquely identifies an attribute. It is formed by concatenating the domain, application, class ID, and attribute name according to specific rules.
Dubbo
Dubbo is a distributed service framework that provides high-performance and transparent remote procedure call (RPC) solutions. It is the core framework of Alibaba's service-oriented architecture (SOA) governance solution. It supports over 3 billion calls for more than 2,000 services daily and is widely used on various Alibaba Group sites.
drm-client
The client JAR package for the dynamic configuration framework.
Dashboard
A dashboard is a page that contains multiple reports.
Single-machine view
Aggregates metric data from the perspective of a single application instance (a single machine). It allows for comparative analysis of application monitoring metrics based on their spatial and temporal distribution on that machine.
Subscription
After subscribing, members of a notification group receive monitoring alert notifications.
Queue
A queue stores messages under a topic. Each topic can have one or more queues. The number of queues for a topic depends on the message type and the region where the instance is located. For the specific number of queues, contact technical support.
Scheduled message
A scheduled message is a message that a producer sends to the message queue server but does not expect to be delivered immediately. Instead, its delivery to a consumer is postponed to a specific time in the future. For more information, see Message types > Scheduled and delayed messages.
Address configuration method
Indicates the source of the system cluster's address.
Unit-based architecture
The application layer uses the same sharding dimension as the data layer to converge the entire request link into a group of servers. This forms a closed unit from the application layer to the data layer. The database only needs to handle requests from the application nodes within its unit, which greatly reduces the number of connections. A unit can be moved as a relatively independent whole. You can even deploy some units to a different location.
F
Fat JAR
A fat JAR is an executable JAR package. It contains compiled classes and the dependent JAR archives required to run the code. You can run the application using the java -jar ***.jar command. A fat JAR differs from a regular JAR because it includes its dependency JAR packages.
FGC
Full Garbage Collection (FGC) is a garbage collection process that covers the entire heap.
Service metrics
Service metrics aggregate and display application-related service metrics, such as Error, Service, SAL, CAL, and DAL. They allow for comparative analysis from spatial and temporal distributions across data centers (DCs), logical data centers (LDCs), and single machines. This provides a single entry point to view and analyze all monitoring data related to an application.
Service mesh
A service mesh is an infrastructure layer that handles inter-service communication. It usually consists of a series of lightweight network agents that are deployed with applications, but the applications do not need to be aware of them. It provides a transparent, programming language-agnostic way to flexibly and easily automate operations such as network configuration, secure configuration, and service observation.
Partitionally ordered message
A partitionally ordered message means that for a specific topic, all messages are partitioned into blocks based on a sharding key. Messages within the same partition are published and consumed in strict First-In, First-Out (FIFO) order. The sharding key is a crucial field in ordered messages used to distinguish different partitions. It is completely different from the message key of a normal message. For more information, see Message types > Ordered messages.
Distributed transaction
A distributed transaction is a transaction where the initiator, resources, resource manager, and transaction coordinator are located on different nodes of different distributed systems.
Branch transaction
A branch transaction, also known as an action, is a part of a distributed transaction. A distributed transaction can contain multiple local database transactions. In a distributed transaction framework, a branch transaction might be an SQL statement executed on a sharded database or a call to a custom-pattern service.
Branch Transaction Records
An Action Record is used to identify a branch transaction. It records information about the participant that provides the branch transaction, including the participant's unique ID. Using the branch transaction information, the transaction manager can commit or roll back the participant's operation.
Initiator
An initiator is responsible for starting a distributed transaction. It incorporates participants into the distributed transaction by invoking their services and decides whether to commit or roll back the entire distributed transaction. A distributed transaction has only one initiator.
Group ID
A globally unique identifier on the gateway used to locate the API being accessed.
Load balancing
The load balancing policy used when there is more than one backend address.
Release window
A release window is an entity that controls and coordinates the unified release of multiple iterations. A release window can be associated with one or more iterations.
G
Gradle
Gradle is a project automation and build tool based on the concepts of Apache Ant and Apache Maven. It uses a Groovy-based domain-specific language (DSL) to declare project settings, which eliminates cumbersome XML-based configurations.
Group
A class of producers or consumers. These producers or consumers typically produce or consume the same class of messages, and their publishing or subscription logic is consistent.
Group ID
The ID of a group.
Workspace
A workspace is a group of resources that are interconnected by a network, have consistent security policies, and have minimal access latency. You can use workspaces to conveniently group and manage resources. For example, you can divide workspaces into development, testing, and production environments based on different development and delivery needs. Resources in different workspaces are isolated from each other. You can assign separate operator permissions to manage each workspace.
Build
A build describes the configuration information for the process of creating an image from application source code. This includes the source code address, branch information, source image access information, destination image information, and Dockerfile location.
Workload
A workload is the carrier of an application in its running state and its higher-level aggregation. It typically includes pods, deployments, StatefulSets, DaemonSets, and jobs.
Work items
Work items are a collection of requirements, bugs, and tasks in a development project.
Fault injection atomic operation
A fault injection atomic operation is the minimum unit of a fault.
Broadcasting consumption
In broadcasting consumption, every consumer identified by a Group ID consumes a specific message once. For example, if a topic has nine messages and a Group ID has three consumer instances, each instance consumes all nine messages in broadcasting consumption mode. For more information, see Clustering and broadcasting consumption.
H
HAS
High Availability Service (HAS) is a high availability management and control platform product focused on disaster recovery. It provides end-to-end disaster recovery capabilities, from customer services to middleware, Platform as a Service (PaaS), and Infrastructure as a Service (IaaS). It supports disaster recovery switchovers and recovery, disaster recovery planning, and disaster recovery drills. It also includes monitoring for the overall status of data centers and disaster recovery, a disaster recovery dashboard, environment inspections, and risk emergency response.
Header parameters
The message header contains several properties in the "Property Name: Property Value" format. The server uses this information to obtain details about the client.
HTTP
Hypertext Transfer Protocol (HTTP).
HTTP method
Indicates the action to be performed on a given HTTP resource.
Golden Signals
The four golden signals are a summary of Google's experience with large-scale distributed monitoring. These four signals help measure user experience, service interruptions, and business impact at the service level. They mainly focus on four types of metrics: throughput, response time, error rate, and saturation. For the service level, the first three metrics are typically used for measurement.
Backend service type
Indicates the type of backend service to which the gateway forwards a request after receiving it.
Backend protocol type
Indicates the type of communication protocol used by the backend service to which the gateway forwards a request.
Backend authentication method
Indicates whether a signature must be added when the gateway sends a request to the backend business system.
I
IAM
The Identity and Access Management (IAM) console of Ant Group is a platform for managing members, assigning permissions, managing identity sources, and viewing operation records.
IDC view
Aggregates metric data from the perspective of a data center (DC). It allows for comparative analysis of application monitoring metrics based on their spatial and temporal distribution on a single machine.
Ingress
Ingress is a collection of rules that authorize inbound connections to reach cluster services. You can configure Ingress to provide externally accessible URLs, load balancing, Secure Sockets Layer (SSL), and name-based virtual hosting.
Istio
Istio is an open source service mesh project. It is a complete, non-intrusive microservice governance solution. Istio provides a simple way to create a network for deployed services with features such as load balancing, inter-service authentication, and monitoring, without any changes to the service code.
J
Jetty
Jetty is an open source Java servlet container. It provides a runtime environment for Java-based web containers, such as JSP and servlets.
Full interface name
The interface name of an RPC interface. It is a class that performs specific functions and is a collection of functionalities.
Image
An image is an application package. It is a binary package that bundles configuration, related software, and other components, and complies with the Docker Image specification. Images can come from any image center accessible to the AKS network. For private image centers, you need to configure the corresponding access information in AKS.
Cluster
A cluster in AKS describes a workload cluster that you create. It consists of multiple nodes.
Clustering consumption
In clustering consumption, all consumers identified by a Group ID share the consumption of messages. For example, if a topic has nine messages and a Group ID has three consumer instances, each instance consumes an average of three messages in clustering consumption mode. For more information, see Clustering and broadcasting consumption.
Cluster task
A cluster task supports splitting data into multiple data chunks in a single execution step. This allows the data of a task to be distributed across different machines for execution. This is suitable for business scenarios with large data volumes. The scheduling process is divided into a splitting phase and an execution phase.
Node
A node is a physical server or virtual machine with Docker and Kubelet installed. It is used to run application workloads.
Monitoring products
Monitoring products are tools provided for implementing custom monitoring. They can be used to configure monitoring metrics and generate data sources.
Monitored object
A monitored object is a target described by one or more dimensions. For example, a container can be described by the single dimension of the container itself. A logical zone of an application needs two dimensions to be described (App + Zone). The monitored object is essentially the owner of the monitoring data.
Monitor mode
In monitor mode, when a traffic-limiting rule is matched, the request is not actually blocked. Only a log of the traffic-limiting record is printed.
Simple job
A simple job is the most basic type of task, suitable for scenarios with simple business logic. A single simple job corresponds to one handler and supports task steps.
Exact match
The request path of the call is fixed to the API request path entered during creation.
Cluster endpoint
The address of the system cluster. It can be an IP address or a domain name.
K
Framework configuration
For applications developed based on the SOFA/SOFABOOT framework, metrics for monitoring items such as PV, Service, SAL, SQL, Cal, and CE Thread are automatically collected based on the default log paths under the framework. You only need to enable monitoring.
L
LDC view
Aggregates metric data from the perspective of a logical data center (LDC). It allows for comparative analysis of application monitoring metrics based on their spatial and temporal distribution on a single machine.
LHC
LDC Hybrid Cloud (LHC) is a unitized application service. It provides capabilities such as application management, release and O&M, traffic steering, and configuration synchronization in multi-cluster Kubernetes scenarios across multiple data centers and regions on top of a cloud-native infrastructure. As a PaaS management and control product that developers and O&M engineers interact with daily, it helps manage applications and logical units, configure and scale units, steer network traffic, and configure monitoring metadata. It not only meets the business needs for multi-active and disaster recovery in financial scenarios but also allows the infrastructure to benefit from containerized infrastructure and cloud-native architecture.
Log4j
Apache Log4j, which stands for "log for Java", is an open source project from Apache. It lets you control the destination of log information and the output format of each log entry. By defining the level of each log message, you can control the log generation process in greater detail.
Log4j 2
Log4j 2 is an upgraded version of Log4j.
Logback
Logback is an open source logging component. SOFABoot uses the SLF4J + Logback logging framework by default.
Traffic failback
The process of switching access traffic back to an application after it has returned to normal operation.
Class ID
A string that represents a configuration class. It is consistent with the ID field of the @DObject annotation in the application code and typically uses the fully qualified class name.
Block Mode
In intercept mode, when a traffic-limiting rule is matched, the request is blocked.
Routing rule
Indicates the routing policy to use after the gateway receives the syntax.
Logical Unit
A unit is called a zone. There are three types of zones: RZone, GZone, and CZone. The characteristics of a zone are as follows:
The same application has independently used resources in each zone.
The services of the same application are horizontally split across different zones.
The service shards processed by different zones do not overlap.
M
Maven
Apache Maven is a project management and build automation tool. It provides developers with a complete build lifecycle framework. Development teams can automatically complete the basic build configuration of a project with minimal effort.
Message ID
A globally unique identifier for a message. It is automatically generated by the message queue system to uniquely identify a specific message.
Message key
A business identifier for a message. It is set by the message producer and uniquely identifies a specific business logic.
Mock
If the backend of an interface is not yet available, you can use a mock to simulate a backend service.
mPaaS mobile application
Represents an mPaaS mobile application. It must be used together with the mPaaS Mobile PaaS.
Namespace
A namespace is consistent with the corresponding concept in Kubernetes. It represents a logically isolated space that isolates elements such as pods, services, and ReplicaSets. However, the network is usually not isolated.
Default second-level domain
Used when a client accesses the gateway using the HTTP protocol. It consists of the group ID and the second-level domain of the gateway.
O
OceanBase
OceanBase is a distributed relational database product independently developed by Alibaba and Ant Group. It combines the advantages of traditional relational databases and distributed systems. It features high availability, high performance, and high scalability. It is compatible with MySQL in terms of functionality and provides finance-grade, highly available database services on general-purpose hardware.
Old Util
Old generation utilization rate.
OSS
Alibaba Cloud Object Storage Service (OSS) is a cloud storage service provided by Alibaba Cloud that offers massive, secure, low-cost, and highly durable storage. Its data durability is not less than 99.9999999999% (twelve 9s), and its service availability (or business continuity) is not less than 99.995%.
OperationType
An API service identifier set for mPaaS mobile applications. It is used to locate the resource to be accessed.
P
POC
Proof of concept (POC).
Producer instance
A producer instance is an object instance of a producer. Different producer instances can run in different processes or on different machines. A producer instance is thread-safe and can be shared among multiple threads within the same process.
Configuration class
A configuration class is a regular Java object in a business application. After being registered with the dynamic configuration framework's programming API, it becomes a resource that can be dynamically managed externally. The domain, application, and class ID uniquely identify a configuration class instance.
Configuration settings
A config map is a native Kubernetes concept used to store user configuration information.
Configuration template
RMS provides a monitoring configuration template feature. It supports exporting application and custom monitoring alerts as a configuration template (JSON file). You can then import it into other environments with one click. If an application with the same name is already deployed in that environment, the monitoring and alerting configuration takes effect immediately. This makes production and O&M more efficient, convenient, and consistent.
Q
Query parameters
Generally refers to the parameters after the ? in a URL.
Frontend protocol type
Refers to the protocol type used when a client sends a request to the gateway.
Request parameters
Indicates the parameters to be configured when a client sends a request to the gateway.
Request body parameters
Refers to the data in the request body.
URI of the request
Indicates the URL of the requested resource. The request path can be used to locate the resource to be requested.
Prefix match
As long as the request path prefixes are the same, they all match this interface, which allows the interface to define multiple different paths.
R
RMC
Resource Management Center (RMC) helps you manage resources by workspace. It does this by binding basic resources, such as Elastic Compute Service (ECS), ApsaraDB RDS, Object Storage Service (OSS), Memcache, and Simple Log Service, to workspaces. This meets the requirements for more granular O&M management in complex, large-scale development and production scenarios.
RMS
Real-time Monitoring Service (RMS) is a finance-grade monitoring product with visualization capabilities. It performs multi-dimensional aggregation on massive amounts of data, including logs, metrics, and traces. It provides users with visualized monitoring functions from multiple perspectives, such as business monitoring, application monitoring, cloud-native monitoring, basic resource monitoring, log query and analysis, and distributed tracing. It has rich visualization dashboards and provides an alert subscription function.
RPC
Remote procedure call (RPC).
RPC ID
The RPC ID is written as RpcId in SOFABoot parameters. The RpcId represents the position of the current request in the entire call chain. For example, if system A calls systems B, C, and D in sequence while processing a request, the RpcIds for these three calls are 0.1, 0.2, and 0.3, respectively. If system B then calls systems E and F, the RpcIds for these two calls are 0.1.1 and 0.1.2.
RPC Service
An RPC service is the implementation object of an interface provided by the server-side.
RPC Reference
An RPC reference is a proxy object created by the client for an RPC service.
RPO
The recovery point objective (RPO) is measured in units of time. It is the point in time to which the system and data must be restored in the event of a disaster. The RPO indicates the maximum amount of data loss that the system can tolerate. The smaller the amount of data loss the system can tolerate, the smaller the RPO value.
RTO
The recovery time objective (RTO) is measured in units of time. It is the time requirement for an information system or business function to be restored after it has stopped due to a disaster. The RTO indicates the maximum amount of time a service can be stopped that the system can tolerate. The more urgent the system service requirements, the smaller the RTO value.
RT distribution graph
A distribution graph of the response time (RT) for requests processed by an application within a specified period. Green dots represent successful request responses. Red dots represent failed request responses.
RPC method
A method in an RPC interface that indicates the specific operation to be performed on a resource.
Pod
A pod is the smallest deployable and manageable unit in Kubernetes. A pod consists of multiple related containers that share a disk.
Toleration
Toleration is the counterpart to the concept of a taint. If pods have a corresponding toleration mark, they can tolerate the taints on a node and be scheduled to that node.
Disaster recovery
Refers to establishing two or more systems with the same functionality in the same city or in different cities that are far apart. The full range of business applications is deployed within the systems. They handle user or other business traffic on a daily basis. When one system fails due to a disaster, natural or man-made, the business system can be quickly switched to another system through business monitoring and function switching. This ensures the continuity of the user's business system to the greatest extent possible.
Switchover
The process of switching traffic to an emergency data center to ensure the normal operation of the business system when a product or application in the production data center fails.
Disaster Recovery
The process of migrating data back to the original data center and restoring the application's operation after the environment in the failed data center has returned to normal.
Disaster recovery plan
An executable plan that includes disaster recovery steps.
Job topology
A collection of tasks triggered by events that describes task dependencies. Subtasks can be executed concurrently or conditionally.
Task orchestration
The process of editing multiple subtasks into a topology task based on business logic.
Task step
A simple job can be broken down into multiple execution steps, which are executed in sequence by different handlers. The next step begins only after the preceding step has been successfully completed.
S
Service ID
A unique identifier for a service. It is composed of the full interface path, version, group, and communication protocol.
Service provider
The application that provides the RPC service.
Service consumer
The application that uses the RPC service.
Service registry
An independent application cluster used to store and maintain a list of all online RPC application addresses.
Service parameters
Parameters of a service provider that can be dynamically modified, such as weight and status.
SOFA
Scalable Open Financial Architecture (SOFA) is a finance-grade distributed middleware framework independently developed by Ant Group.
SOFABoot
A lightweight middleware integration solution based on Spring Boot. It integrates seamlessly with standard Spring Boot projects and includes a full set of SOFA finance-grade middleware.
SOFAStack Service Mesh
A large-scale implementation of Service Mesh based on improvements and extensions to Istio. It can meet the performance requirements of large-scale deployments and handle practical situations in implementation.
SOFATracer
SOFATracer is a component for distributed system call tracing. It uses a unified traceId to log various network call situations in the call chain, providing visibility into network calls. These logs can be used for rapid fault discovery, service governance, and more.
SOFAMosn
SOFAMosn, which stands for Modular Observable Smart Network, can be used as a data plane sidecar in SOFAMesh. It is written in Go and is compatible with Envoy's API, allowing it to integrate with Istio.
SOFAREST
SOFAREST is a Service-Oriented Architecture (SOA) solution based on the Java API for RESTful Web Services (JAX-RS) standard.
SOFARPC
An RPC framework independently developed by Ant Group. SOFARPC provides point-to-point service invocation functionality between applications.
Spring Boot
Spring Boot is a framework used to simplify the initial setup and development process of Spring applications. The framework uses a specific configuration method, so developers no longer need to define boilerplate configurations.
Spring Cloud
Spring Cloud is a collection of frameworks that uses Spring Boot to simplify the development of distributed system infrastructure. Features such as service discovery and registration, configuration centers, message buses, load balancing, circuit breakers, and data monitoring can all be started and deployed with one click in the Spring Boot development style.
Starter
A starter for Spring Boot/SOFABoot that allows for quick integration of embedded functional modules.
SRE
A Site Reliability Engineer (SRE) is a combination of a software engineer and a system administrator.
SSL
Secure Sockets Layer (SSL) is a standard protocol used to encrypt communication between a browser and a server.
Data scrape latency
Data scrape latency is the time difference between when the data synchronization service scrapes data and writes it to the product's storage, and when the data is actually written to the source database. Data scrape latency only has a value during the incremental synchronization phase.
Data write latency
Data write latency is the difference between the timestamp of the latest data synchronized to the destination instance as executed in the source database, and the current timestamp of the source instance. Synchronization latency reflects the time difference between the data in the destination instance and the source instance. When the synchronization latency is 0, it means the data in the source and destination instances are completely consistent. Data write latency only has a value during the incremental synchronization phase.
Data chunk
When the data volume is large, the task scheduling system supports splitting the data into multiple chunks in a single execution step. This distributes the data of a task across different machines for execution. The specific scheduling method is adjusted by the scheduling center based on the load progress.
Application
The application to which the configuration class belongs.
Attribute
A private attribute of a configuration class object that has public read and write methods. A configuration class can have multiple attributes. One configuration class attribute corresponds to one configuration item of the business.
Attribute name
A string that represents an attribute, consistent with the naming of the private attribute in the business code.
Transaction
A transaction is a series of operations performed as a single logical unit of work, which are either fully executed or not executed at all.
Transactional message
Message Queue provides a distributed transaction function similar to X/Open XA. Transactional messages from Message Queue can achieve eventual consistency in distributed transactions. For more information, see Message types > Transactional messages.
Transaction manager
The transaction manager is an independent service that coordinates distributed transactions. This includes creating main transaction records and branch transaction records, and invoking participant commit or rollback methods based on the state of the distributed transaction.
Ordered message
A type of message provided by Message Queue that is published and consumed in order. It is divided into globally ordered messages and partitionally ordered messages. Currently, only partitionally ordered messages are supported. For more information, see Message types > Ordered messages.
T
Tengine
A web server project initiated by Taobao. It is based on Nginx and adds many advanced features and characteristics for high-traffic websites.
Tomcat
Primarily serves as a Java servlet/JSP container, but also has many features of a traditional web server.
TraceId
A TraceId refers to the ID in a tracer that represents a unique request. This ID is generally generated by the first system in the cluster to process the request.
Tracer
Identifies the entire request chain, which is a combination of a series of spans. Its own ID is carried throughout the call chain, and each span within it must carry this traceId. Therefore, the traceId is passed throughout the call chain.
Notificant
The recipient of an alert message, who receives it by text message.
Notification group
Each notification group can contain one or more notificants. When managing notifications, you can subscribe through a notification group to send alert content to all notificants within that group.
Throughput
The throughput trend of requests processed by an application within a specified period.
W
Taint
A taint is a native Kubernetes concept used to mark nodes. It is typically used for pod scheduling policies.
Physical table ID
A globally unique ID for a physical table.
X
Response parameters
Used to interpret the content of the response.
Response body parameters
Used to explain the meaning of parameters in the response body.
Response example
Used to show an example of the response status.
Post-throttling operation: null process
No action is taken, and the process returns directly. For an interface method, it returns null. For a web page, it returns empty and ends the current page visit.
Message
The carrier of information transmission in a message queue.
Topic
A first-level message type used to classify messages. For more information, see Topics and tags.
Message tag
A second-level message type used to further classify messages under a specific topic. For more information, see Topics and tags.
Producer
Also known as a message publisher, responsible for producing and sending messages.
Consumer
A consumer, also known as a subscriber, receives and consumes messages.
Message accumulation
A state where a producer has sent messages to the message queue server, but due to limited consumer capacity, not all messages are consumed correctly in a short period. The unconsumed messages are stored on the message queue server.
Message filtering
A consumer can filter messages based on message tags to ensure that it only receives the filtered message types. Message filtering is completed on the message queue server. For more information, see Message filtering.
Message trace
The complete link information formed by data such as time and location from various relevant nodes during the process of a message being sent from a producer to being consumed by a consumer. With a message trace, you can clearly locate the complete link of a message from the producer, through the message queue server, to the consumer, which is convenient for locating and troubleshooting problems. For more information, see Query message traces.
System cluster
Represents the cluster of the real business system to which the gateway forwards frontend requests.
Project
A project has a unique process with a start and end date. It consists of a series of coordinated and controlled activities. The process is implemented to achieve specified goals, including meeting constraints such as time, cost, and resources.
Y
Business error code
Used to explain the meaning of business error codes in the response.
YGC
Young Generation Garbage Collection (YGC).
Application
Refers to the applications that make up a business system. It can be a monolithic application or a microservice application based on a distributed framework.
Application topology
A topology is a visual representation of the call relationships and dependencies between applications.
Application service
This concept is consistent with the concept of an application service in classic application services. However, because containers have their own special characteristics, application services in AKS contain some additional metadata, such as container specification configuration, images, scheduling policies, and log configurations.
Application container
The isolated workspace where an application runs. It is usually a Docker container or a sandbox workspace with isolation capabilities that is compatible with the CRI interface, such as Pouch.
Inplace upgrade
Inplace upgrade refers to the update method for pods in an application service. In this method, the IP address of the pod remains the same after the release as it was before, and the node it is on remains unchanged as much as possible.
Cloud-native deployment
Cloud-native deployment also refers to the update method for pods in an application service. In this method, the IP address of the pod is usually different after the release than it was before, and the node it is on may also change.
Page liveness probe
Refers to determining the liveness of a page by opening the inspection page in a browser. In addition to supporting static pages that do not require logon, the High Availability Service platform also supports liveness probes for pages that require a logon state. It also supports matching the content or elements of a page to determine if the page has rendered successfully. On the High Availability Service platform, you can configure page liveness probes as inspection tasks to periodically inspect pages.
Alert
Refers to triggering an alert notification through configured alert rules.
Yunyou
Yunyou is Ant Group's one-stop Apsara Stack planning, delivery, and O&M platform. It manages the entire lifecycle of Apsara Stack from its inception to its landing.
Domain
A namespace for a configuration class. The default value is Alipay and can be modified through programming annotations.
Running mode
Refers to how the traffic-limiting and circuit-breaking client handles traffic limiting. It is divided into monitor mode and intercept mode.
Delayed message
A delayed message is a message that a producer sends to the message queue server but does not expect to be delivered immediately. Instead, its delivery to a consumer is delayed for a certain period. For more information, see Message types > Scheduled and delayed messages.
Iteration
An iteration is a development activity to implement a requirement or fix a bug. An iteration can include one or more applications.
Metadata info
A collection of an application's basic information and development configurations.
Z
Resource
Refers to environment isolation services, computing, network, storage, and database resources. Examples include Workspace, Elastic Compute Service (ECS), VPC, Server Load Balancer (SLB), Object Storage Service (OSS), and ApsaraDB RDS.
Metric
A concept and value that quantifies a characteristic of a monitored object. You can use several metrics to describe how a monitored object changes over time. This helps you monitor what happened to which object.
Fund security monitoring
A risk control product that helps you find the risk of financial loss. The product focuses on monitoring financial data. It executes user-defined checkpoints based on various verification policies, such as daily and hourly, to find financial data issues and promptly send you alerts. You can receive alerts right away, troubleshoot problems based on the abnormal data, analyze the causes, and then fix vulnerabilities to avoid financial loss.
Custom monitoring configuration
For applications not developed with the SOFA/SOFABOOT framework, you can configure the log collection path and column values for each metric, such as Error and Dal. For applications developed with the SOFA/SOFABOOT framework, if you do not want to use the default log path for monitoring, you can also switch to Custom monitoring configuration.
Primary transaction record
An Activity Record is the main body of the entire distributed transaction. Its core data structure consists of the transaction ID (TX_ID) and the transaction state (STATE). It is persistently written to the database when the distributed transaction is started. Its state determines the state of the distributed transaction.