Operational Overview

Overview

Operational Overview provides visualization and multi-dimensional statistics for the Fastone platform, helping operations teams understand overall platform usage and health.

Operational Overview vs Monitoring

	Operational Overview	Monitoring
Scope	Platform, cluster, task, and user dimensions	Host-level resource utilization and service runtime status
Time	Long-term historical analysis	Short-term monitoring
Availability	Advanced feature (for example FCC-E Advanced/Enterprise editions)	Basic feature (all editions)
Permissions	Platform operations focused, admin-only	Available to all users

Permissions

• Only administrators can access the Operational Overview submenu. It is not available to regular users.

Time range selector

The time range selector applies to platform, cluster, task, and user dimensions. Changing the time range updates all time-based charts. You can select a preset past period or a custom time range.

Note: For "past N days", today is not included. For example, if today is 2022-10-03 and you select "past 2 days", the range is 2022-10-01 00:00:00 to 2022-10-02 23:59:59.

Platform dimension

• Tasks: Total number of tasks in all states.
• Clusters: Total number of clusters in all states.
• Users: Total number of users in all states.
• Nodes: Total number of desktop/cluster nodes in all states.
• CPU cores: Sum of CPU cores across running cluster/desktop nodes.
• Total memory: Sum of memory across running cluster/desktop nodes.

Node status counts at the current time.

Based on the time range, shows the number of nodes created in the period and their current status distribution.

Based on the time range, shows line charts of status count changes over time.

Cluster dimension

Shows allocated CPU, allocated memory, CPU utilization, and memory utilization for available clusters at the current time.

Based on the time range, shows line charts of CPU and memory utilization for available cluster nodes.

Cluster status counts at the current time. Based on the time range, shows the number of clusters created in the period and their current status distribution.

For the selected cluster and time range, shows changes in the number of running nodes, including head nodes, login/submit nodes, and compute nodes.

Based on the selected time range (filtered by job start time), shows job details for the selected cluster.

Task dimension

Task status and task count

Task status: shows a task status distribution chart.

Task count: based on the selected time range, shows counts by status.

Application usage (task mode)

Based on the task list, groups by application and sorts descending. Shows total usage (core-hours) per application and total usage across all applications in the selected time range.

The top application is shown as 100% and others are scaled relative to it.

Task list

Filters jobs by task creation time based on the time range selector.

The task list includes fields such as:

• Task name, cluster, application, task ID, status, created/start/end time, total duration, usage (core-hours), username, selected cluster, instance type, hyper-threading, system volume size, application parameters, log level, remove intermediate results.
• Usage (core-hours): Core-hours requested by all jobs submitted for the task, calculated as requested CPU cores x runtime hours. It may be empty if the task has not completed. Example:
  • A task submitted via raw-command typically submits one scheduler job, so usage = requested CPU cores x runtime.
  • A task submitted via vina may submit multiple jobs, so usage = sum(requested CPU cores x runtime) across jobs.
  • Example: with a 2-core cluster, running sleep 3600 for 1 hour results in usage = 2 x 1 = 2.00 core-hours.

Functions:

• Search and sort: supports searching/sorting by task name, cluster, application, status, created time, total duration, username, selected cluster.
• Export: export filtered data to CSV using selected columns.
• Custom columns: select which columns to display.

User dimension

Daily active rate

Based on the selected time range, users are sorted by active days descending and shows active days and active rate.

Definition:

If a user makes any Fastone API request on a day, the user is considered active on that day.

Examples:

• Past 7 days: active days range is 0 to 7. Users with 0 active days are not shown. If a user has 5 active days, active rate is 5/7 = 71.4% (rounded to 2 decimals).
• Past month: active days range is 0 to 30. Users with 0 active days are not shown. If a user has 10 active days, active rate is 10/30 = 33.3% (rounded to 2 decimals).

User usage (task mode)

Based on the task list, groups by user and sums usage (core-hours) per user within the selected time range.

User usage (cluster mode)

Cluster selection defaults to All. You can also select a specific Fsched cluster (including clusters in all states).

If a specific cluster is selected, usage is computed from that cluster's job list, grouped by user and filtered by job start time within the selected time range.

If All is selected, usage is computed from job lists across all Fsched clusters, grouped by user and filtered by job start time within the selected time range.

Node fields (reference)

• System disk: Current system disk capacity.
• Connection status: Connected/Disconnected, whether a session is connected to the node.
• Created at: Node creation time.

Actions:

• Power on/off: same as cluster node power operations.
• Image: same as cluster node image operations.
• Edit: same as cluster node edit operations (edit host name and instance type).
• Release: same as cluster node release operations.
• Bulk actions: same as cluster node bulk actions.