Resources Allocations
This page explains how computational resources and allocations are defined and used within CCS systems.
Resources
A resource is a physical or virtual computing system that provides compute capability or storage capacity.
Examples of CCS resources include:
CPU compute nodes
GPU compute nodes
OpenStack virtual machines
Parallel file systems (for example, GPFS storage)
Object storage systems (for example, Ceph storage)
Network-attached storage (NAS)
Block storage devices
There may be multiple instances of a resource type.
For example:
PKS2 Ceph Object Storage
CoT Ceph Object Storage
Each instance is managed independently and may have different performance characteristics or access policies.
Allocations
An allocation is a defined amount of compute time or storage space granted on a specific resource instance.
Allocations are typically assigned to Principal Investigators (PIs) and are used to support research, teaching, or infrastructure development.
Allocation Types
CCS supports several allocation categories.
Educational Allocations
Resources provided to support instructional activities such as coursework or training programs.
CCS Discretionary Allocations
Resources managed by CCS to support urgent, exploratory, or unexpected computing needs.
Condo Allocations
Allocations associated with hardware or storage resources purchased by research groups and contributed to shared CCS infrastructure.
Open Access Allocations
Shared resources made available to all PIs when capacity exists.
These allocations represent:
Unused compute time on shared systems
Access to temporary storage areas such as scratch space
Scratch storage is not permanent and data may be removed after a defined retention period (for example, 90 days).
Purchased Resources
Research groups may purchase dedicated hardware (such as condo compute nodes) or storage capacity.
These purchases result in:
Condo allocations
Condo incentive allocations
These allocations provide priority or guaranteed access to the purchased resource share.