{% extends "base.html" %} {% import 'utils.jinja' as utils %} {% block style %} {% endblock style %} {% block js %} {{ super() }} {% endblock js %} {% block body %}
Contribution of compute capacity from the
{{ ce_facility_name }} entity named {{ ce_site_name }}.
Description of this Compute Entrypoint (CE): {{ ce_description }}
Displays the CPUs contributed by this site and allocated to project jobs.
Displays the memory contributed by this site and allocated to project jobs.
Displays the disk contributed by this site and allocated to project jobs.
Displays the Basic Compute Units (BCUs) contributed by this site and allocated to project jobs. 1 BCU is defined as 1 CPU core, 4 GB of memory, and 4 GB of disk. Total BCUs are calculated based on the most constrained resource.
For example, if a site contributes the following:
Its BCU contribution would be 2 BCUs, since the CPU cores are the limiting factor.
{% endcall %} {% call utils.bootstrap_modal('memory-per-cpu-modal', 'About: Memory per CPU') %}Displays the ratio between memory and cpus contributed by this site and allocated to project jobs.
The difference between the Contributed and Allocated Memory per CPU is an indicator of the efficiency of resource utilization at a site. When the allocated ratio is larger than the contributed ratio, it suggests that jobs are using more memory per CPU than what was provisioned, and may become memory starved. When the allocated ratio is smaller, it indicates that jobs are using less memory per CPU than what was provisioned, and may become CPU starved.
{% endcall %} {% call utils.bootstrap_modal('disk-per-cpu-modal', 'About: Disk per CPU') %}Displays the ratio between disk and cpus contributed by this site and allocated to project jobs.
The difference between the Contributed and Allocated Disk per CPU is an indicator of the efficiency of resource utilization at a site. When the allocated ratio is larger than the contributed ratio, it suggests that jobs are using more disk per CPU than what was provisioned, and may become disk starved. When the allocated ratio is smaller, it indicates that jobs are using less disk per CPU than what was provisioned, and may become CPU starved.
{% endcall %} {% call utils.bootstrap_modal('groups-per-day-modal', 'About: Groups Served per Day') %}Displays the count of different organizational units that have utilized resources at this site.
{% endcall %} {% call utils.bootstrap_modal('resources-per-day-modal', 'About: CPU (l), Memory (r), Disk (r) Allocated per Day') %}Displays the sum total of resources contributed by the site per day, multiplied by the duration for which they were contributed. For example, if 4 CPUs were allocated for 5 hours, that would contribute 20 CPU Hours to the total for that day.
{% endcall %} {% call utils.bootstrap_modal('resource-limited-per-day-modal', 'About: Maximum Concurrent Resource-Limited Glideins Per Day') %}For each resource type, display the maximum percentage of running Execution Points (EPs) that were limited by that resource on a given day. Here, "limited" is defined as having over 90% of a resource contributed to the EP allocated to running jobs.
For example, if an EP has 16 GB of memory and jobs running on the EP occupy 15 GB of memory, then the EP would be considered memory-limited. If 8 of the 10 EPs running at the site were memory-limited, then the site would be 80% memory-limited at that time.
Values displayed are the maximum percentages observed on that day. If, at any point in the day, 90% of concurrently running EPs were CPU-limited, the value for that day would be 90%.
{% endcall %} {% call utils.bootstrap_modal('eps-past-week-modal', 'About: Running Execution Points Past Week') %}Displays the count of running execution points at this site.
{% endcall %} {% call utils.bootstrap_modal('allocation-modal', 'About: CPU Allocation Percentage by Project') %}For the given time range (day, week, or month), display the percentage of total CPU allocation at the site used by each project that ran jobs at the site.
{% endcall %}