Monitoring: core components

For the quick access to core nodes and services, a new tab was added to the Cluster State section - CORE NODES tab:

This tab includes two subtabs:

  • Core nodes - to view the list of all core nodes of the current platform deployment
  • Core services - to view the list of all services on the core nodes and their state

Please note, CORE NODES tab is available and may be viewed only by admins.

Core nodes

To view the list of platform core nodes:

  • open the Cluster State section
  • navigate to the CORE NODES tab → Core nodes subtab

This subtab shows the table similar to the one for general active nodes (ALL NODES tab of the Cluster State section), but only for core nodes.

Displaying details:

  • Name - name of the node
  • Pipeline - currently assigned run on the node (usually not applicable)
  • Labels - characteristics extracted from the parameters of the node. There are common labels:
    • node name
    • labels with details of compute instances used for nodes
    • labels of core services used by nodes
  • Addresses - node addresses
  • Created - date of the node creation

To get currently active nodes list, use the !button in the right-upper corner of the nodes list.

To navigate to the detailed node information page, click its row in the table.

Core nodes table supports sorting by any of columns:

  • Name
  • Labels
  • Created

To sort the table, click the sorting control near the column header.

Core nodes table supports filtering by any of columns:

  • Labels
  • Addresses

To filter the table:

  • click the filter control near the column header:

  • in the appeared filter, specify desired value and click the OK button, e.g.:

  • table will be filtered:

Core services

To inspect the list of services of platform core nodes:

  • open the Cluster State section
  • navigate to the CORE NODES tab → Core services subtab

This subtab shows the list of services for core nodes.

Each service presents an entity that can be expanded - via the plus button (!) near the service name.

Inside the service entity, there is a list of service pods. Each pod entity can be expanded as well - via the plus button (!) near the pod name.

Inside the pod entity, there is a list of pod containers.

To collapse the entity, use the minus button () near the entity name.

To collapse all expanded entities, use the corresponding button in the right-upper corner of the table.

To get the current services list, use the !button in the right-upper corner of the table.

Displaying details in the table:

  • Name - name of the service/pod/container
  • Node name - name of the pod's node
  • Namespace - pod namespace
  • Pod status - state of the service/pod/container
  • Restarts - number of pod restarts
  • Uptime - duration of the pod uptime

Service details

You may view detailed info of a pod:

  1. Expand the desired service.
  2. Click the info icon near the pod name, e.g.:
  3. Pop-up with the pod info in JSON format will appear:
  4. For some pods, that pop-up can include additional tab - Events, e.g.:
  5. This tab includes pod events list, e.g.:

You may view logs of a container:

  1. Expand the desired service.
  2. Expand the pod.
  3. Click the info icon near the container name, e.g.:
  4. Pop-up with the container running logs will appear:
  5. You may view/refresh such logs in the pop-up or download them to the local workstation via the corresponding buttons.

Service status

Status shown in the Pod status column varies by the entity type.

For services, status has the format [X]/[Y], where:

  • [X] - number of not failed pods of the service. Note: not failed pods may be running, succeeded or pending.
  • [Y] - total number of the service pods

Additionally, there is a colorful icon that shows service status:

  • Unhealthy services (i.e. services with at least one failed pod) has a red status circle
  • Healthy services (i.e. services without failed pods) has a green status circle

For pods, status may be one of the following:

  • Pending, status icon is orange
  • Running, status icon is blue
  • Succeeded, status icon is green
  • Failed, status icon is red

Containers have statuses from the same categories as pods, with the corresponding icons. Name of these statuses may vary from the category names depending on the container state.

Filters

There are several ways to filter core services:

  • Main filter - allows to show only healthy / unhealthy / all core services - this filter is located above the table and presents the corresponding control, e.g.:

  • Column filter - allows to filter table by the specific value(s) in the desired column(s):

    • click the filter icon in the column header
    • specify/select the desired value in the filter field
    • table will be automatically filtered by the specified value, e.g.:

  • Search field - allows to perform the search over the whole services table, e.g.:

To reset all specified filters, click the button Clear filters in the top of the table.