Terminology¶
This section introduces the major concepts involved in EnOS™ ECP.
OU¶
Tenant of the enterprise container platform, owner of the resources.
Project¶
More fine-grained people and resource management units under the same OU. Projects in the developer studio can correspond to projects in the namespace, code repository, and image repository managed by the container cluster, thus enabling resource isolation at the project level. In cluster management, multiple virtual clusters are supported, which depend on the same physical cluster at the bottom. These virtual clusters are called namespaces.
Application¶
Multiple applications can be managed under the same project. Applications are divided into the following categories, depending on the development process and operations.
Normal
Repository
Mobile
Chart
Function
Container¶
A container is a special process model that provides hardware, operating system, storage, network capacity, and other resources required for application deployment and operation, and provides high-performance and scalable management services for application development.
Cluster¶
A cluster is a collection of computing, storage, and network resources that can be used to run various container-based applications.
Pipeline¶
A pipeline is the complete process from application development to delivery that connects the different phases and tasks (build, code scanning, Jenkins jobs, etc.) in the process and can be set to execute step by step automatically.
Job¶
A Job is a resource object used in the Enterprise Container Platform to control batch-type tasks, i.e., jobs that are executed only once and that guarantee the successful completion of one or more Pods of a batch task. A Pod managed by a job is automatically exited when the task completes successfully according to the user’s settings. For example, executing a job to upload an image to a mirror repository before creating a workload.
Pod¶
The smallest unit of the Enterprise Container Platform for resource scheduling, with one or more closely related business containers running in each Pod. See Viewing Pod Information.
Node¶
The compute power in an Enterprise Container Platform cluster is provided by Node, which is the working host where all Pods run, either physical or virtual machines.
Event¶
Kubernetes (K8s) Event is a resource type in K8s that automatically creates and logs things when other resources have state changes, errors. In the Enterprise Container Platform, it can be viewed in the Events tab on the details page for different types of resources in the application and K8s.
The event key message contains:
Reason：A human-readable description of the status of this operation.
Message：This should be a short, machine understandable string that gives the reason for the transition into the object’s current status.
involvedObject：The object that this event is about.
Type：There are currently only two types of events in K8s: ‘Normal’ and ‘Warning’.
Persistent Volume¶
Persistent Volume (PV) is a storage resource within a cluster, similar to how a node is a cluster resource. PVs are independent of the Pod lifecycle and can be created with different types of PVs depending on the StorageClass type.
Persistent Volume Claim¶
Persistent Volume Claim (PVC) is the user of the resource. Like Pods consume node resources, PVCs consume PV resources.
Ephemeral Storage¶
Pods ephemeral local storage can often be used to implement functions such as buffering, log saving, etc.
Configure ephemeral storage limit in the project: Project Management Quota.
Configure ephemeral storage limit in Pod: Resource quota for Deployment.
Note
When the ephemeral storage limit is exceeded, the Pod will be evicted.
For more information on ephemeral storage, see Kubernetes Website。
Deployment¶
Stateless workloads (Deployment), stateless workloads represent a single update operation to a Kubernetes cluster. It is suitable for scenarios that run completely independent, functionally identical applications. See Configuring Deployments.
StatefulSet¶
StatefulSet supports orderly application deployment, scaling, and rolling upgrades as they occur. If you wish to use Persistent Volume to provide persistent storage for workloads, you can use StatefulSet as part of your solution. See Configuring Stateful Sets.
Service¶
An Enterprise Container Platform service is a minimal object, similar to a Pod, and like other endpoint objects. See Configuring Services and Routes.
ConfigMap¶
ConfigMap can be used to store fine-grained information such as individual properties, or coarse-grained information such as entire configuration files or JSON objects. You can use ConfigMap to store configuration information and configuration files that do not require encryption. See Configuring Config Maps and Secrets.
Secret¶
Secrets are used to store information about the use of some sensitive configurations in a Kubernetes cluster, such as passwords, certificates, etc. See Configuring Config Maps and Secrets.
Storage¶
Storage, which provides the administrator with a configuration method for describing the storage ‘Class’. See Configuring Storage.
Blue/Green Deployment¶
Provides a zero-downtime deployment method, deploying the new version while keeping the old one, putting both versions online at the same time, so that if there are problems, they can be handled quickly. See Publishing an Application by Blue/Green Deployment.
Release Records¶
Instances generated by the same Chart deployment under the same project become Release records. Users can view all the release records under the selected project in the App Hub > Releases menu.
Different versions of the same Chart can be continuously updated on the same release record, and Helm will check the differences between the original version and the existing version to overwrite each update.
If you need to deploy multiple instances of a Chart under a project, you need to enter different release names when deploying.