Collabora Online resource allocation and sizing guide
This guide provides hardware allocation and container deployment sizes for Collabora Online.
Note
The resource metrics, ratios, and formulas presented in this guide are based on general guidance and benchmarks provided by Collabora. For the latest and authoritative sizing guidance, see Hardware Provisioning and Collabora Online Technical Information (PDF).
Concurrent user sizing quick reference
The table below provides high-level, rounded planning values for peak Active Concurrent Users (CUs). Validate final production sizing against the latest Collabora guidance in the linked SDK and PDF documentation.
| Active Concurrent Users (CUs) | vCPU Required | RAM Required | Network Bandwidth |
|---|---|---|---|
| 50 CUs | 4 vCPU | 4 GB | 5 Mbit/s |
| 100 CUs | 7 vCPU | 6 GB | 10 Mbit/s |
| 250 CUs | 17 vCPU | 14 GB | 25 Mbit/s |
| 500 CUs | 34 vCPU | 26 GB | 50 Mbit/s |
| 1,000 CUs | 67 vCPU | 51 GB | 100 Mbit/s |
Container deployment profiles
When deploying Collabora Online in containerized environments, resources can be allocated using standard Profiles. Use these profile values as a quick starting point, then tune based on workload behavior and the latest Collabora recommendations:
| Profile | Active Capacity Target | vCPU Allocation | RAM Request / Limit | Local Disk Storage |
|---|---|---|---|---|
| Small Profile | \~30 Active Concurrent Users | 2 vCPU | 3 GB / 4 GB | 10 GB |
| Medium Profile | \~60 Active Concurrent Users | 4 vCPU | 5 GB / 8 GB | 20 GB |
| Large Profile | \~120 Active Concurrent Users | 8 vCPU | 10 GB / 16 GB | 40 GB |
Key allocation guidelines
- Memory Limits: Setting RAM limits higher than initial requests provides necessary safety headroom when users open unusually large or complex spreadsheets and presentations.
- Profile Selection: The Medium Profile is generally recommended as a balanced starting point for most worker nodes, whether running on bare metal or virtualized infrastructure.