Services pages

The overview page for services displays the tenant’s services and applications:

• It shows the status of the “CPU” and “Memory” resources.
• Using the buttons at the top, you can do the following:
  • + Flink Cluster: Create an Apache Flink cluster
  • + Kafka Proxy: Create a Kafka Proxy service
  • App Catalog: Navigate to the App Catalog, where you can deploy pre-packaged applications
  • + Service: Add your own custom service, based on a container image in the DSH’s Harbor Container Image Registry. See Custom service for more information.
• The page itself displays the tenant’s services and applications, with information about their version, status, health, etc.

CPU and memory

Your services and applications are closely linked with the “CPU” and “Memory” resources. If your tenant doesn’t have enough of these resources available, then your services or applications fail to start or restart. For that reason, the overview page for services displays the current status of the “CPU” and “Memory” resources:

• The horizontal bars and the metrics refer to the limit and the allocated CPUs or memory.
• Click the “+” icon next to the horizontal bars to request a new “CPU” limit or “Memory” limit.

Tip

Visit the following pages for more information about how CPU and memory can influence your services or applications:

• CPU limits
• CPU in the service definition
• Memory limits
• Memory in the service definition

Overview of services

The overview page lists the different types of applications or services. The overview of the services has the following logic:

• Every custom service corresponds to one single service in the list.
• An application from the App Catalog contains one or more services.
• The Kafka Proxy and Apache Flink also contain one or more services.

The icons at the left of each line refer to the type of application or service:

Icon	Type of application or service
	A custom service, based on a container image that the tenant uploaded. See Custom service for more information.
	An application from the App Catalog. See App Catalog for more information.
	A single service under an application, the Kafka Proxy, or Apache Flink.
	The "Kafka Proxy" service. See Kafka Proxy for more information.
	The "Apache Flink" service. See Stream processing for more information.

For every service, the DSH displays the following information:

• Name: The name that the tenant chose for the application or service
• Version: The version number of the application or service
• Instances: The number of running instances for the service
• CPU: The number of CPU cores for the service, as defined in its configuration or service definition
• Memory: The amount of memory for the service, as defined in its configuration or service definition
• Status: Icons that indicate the status of the service:
  • A spinner indicates that the DSH is deploying the service.
  • A checkmark indicates that the service is running.
  • A cross indicates that the service has stopped.
  • A blue flag indicates that there are informative notifications about the service.
  • A red exclamation mark indicates that there are errors for the service.
• Health: A heart pulse icon that indicates the health of the service, based on the parameters such as service health check, CPU, memory, or consumer group lag:
  • A green icon indicates that all parameters are well within their limits.
  • An orange icon indicates that one or more parameters are going near their limits. For example, a task is using more than 90 % of its allocated memory.
  • A red icon indicates that at least one parameter violated its limit. For example, a task is using more CPU than was allocated.
  • A black icon indicates that the service isn’t running.

Managing services and applications

On the overview page, you can click the line of an application or service to inspect its details:

• The details page of the Kafka Proxy, Apache Flink or of an app from the App Catalog displays a description, the current configuration, and related services and resources.
• The details page of a service or a custom service displays the health status and tasks.

Apache Flink

The details page of Apache Flink displays the following information:

• In the top right corner of the page, a button to destroy the instance of Apache Flink, and a button to navigate to the interface of Apache Flink
• On the left, the current configuration
• On the right, an overview of the services (“applications”) and resources that Apache Flink uses

Click the services under “applications” to navigate to their details page. See Service for more information.

App from App Catalog

The details page of an app from the App Catalog contains the following information:

• In the top right corner of the page, buttons to start, stop or remove the app, and to edit its configuration. See Buttons for more information about the buttons and the actions they represent.
• On the left, a description of the app
• Under “Services & resources” on the right, a list of the services and resources that the app uses
• Under “Configuration” on the right, an overview of the current configuration of the app

Click a service to navigate to its details page. See Service for more information.

Kafka Proxy

The details page of the Kafka Proxy displays the following information:

• In the top right corner of the page, a button to destroy the Kafka Proxy
• On the left, the current configuration
• On the right, an overview of the services (“applications”) and vhosts that the Kafka Proxy uses

Click the service under “applications” to navigate to its details page. See Service for more information.

Service

The details page of a service displays the following information:

• In the top right corner of the page, buttons to manage the service:
  • For a custom service, you can use these buttons to edit the service definition, and to start, scale, stop or remove the service. See Buttons for more information about the buttons and the actions they represent.
  • For platform services, you see one single button that leads you to the application (App Catalog, Kafka Proxy, Apache Flink) that uses the service.
• Below the title, you see the number of instances, the status and the health of the service.
• Under “Running tasks”, you see a list of the running tasks for the service, and the following information:
  • ID: The unique identifier of the task
  • Staged at: The date and time when the DSH staged the task
  • Started at: The date and time when the task started
  • CPU: The number of CPU cores that the task currently uses
  • MEM: The amount of memory, in megabyte (MB), that the task currently uses
  • RX: The amount of data, in bytes, that the task received from the DSH’s network
  • TX: The amount of data, in bytes, that task transmitted on the DSH’s network
  • Status: The current status of the task in question:
    • Staging: The DSH is making preparations to start the task, for example by creating space to deploy the container.
    • Starting: The DSH is starting the task.
    • Running: The task is currently running on the DSH.
    • Killing: The DSH is killing the task, for example because a user stopped the parent service in the DSH Console.
    • Killed: The DSH killed the task, for example because a user stopped the parent service in the DSH Console.
    • Failed: The task failed to start.
    • Finished: The task finished running. Note that the DSH assumes that all tasks keep running until a user or the DSH kills them. For that reason, the DSH associates a task that finishes automatically with an error.
    • image_error: The DSH encountered an error with the container image when it started the task. Make sure that service definition contains the correct reference to the image in the DSH’s Harbor container image repository, and that this reference doesn’t contain a typo or a non-existent version.
    • oom_killed: The DSH killed the task because it went over the assigned memory limit. This might indicate that you set the memory limit too low, or that the task has a memory leak.
    • Unknown: The DSH can’t retrieve the current status of the task.
  • Health: A button to navigate to the health page of the task
  • Logs: A button to navigate to the Loki logs of the task
• Under “Stopped tasks”, you see a list of the stopped tasks for the service, and the following information:
  • ID: The unique identifier of the service
  • Staged at: The date and time when the DSH staged the task
  • Started at: The date and time when the task started
  • Stopped at: The date and time when the task stopped
  • Status: The status of the task
  • Logs: A button to navigate to the Loki logs of the task
• Under “Health of service”, you see 3 parameters that the DSH uses to calculate the health of the service:
  • CPU: The total number of CPU cores that all the running tasks currently use, and the number of allocated cores
  • Memory: The total amount of memory that all the running tasks currently use, and the amount of allocated memory
  • Consumer group lag: The difference between the offset committed by the consumer group, and the offset of the latest published message in the topic partition. It indicates how far behind the consumer group is in processing the messages in that topic partition. See Consumer groups for more information.
  • A green heart pulse icon () indicates that the parameter is well within its limits.
  • A red heart pulse icon () indicates that the parameter violates its limit.
  • A black heart pulse icon () indicates that the parameter isn’t relevant for the task, for example because it doesn’t use a Kafka consumer group.