Browse Workloads
Browse deployed Pod workloads and view their details in AI Inference.
Prerequisites
- Account with AI Inference access
- A cluster and project selected
Step 1. Go to Workloads
- Click Infrastructure → Workloads in the left sidebar.
- The Pod dashboard for the current project is displayed.
Step 2. Search and Filter
| Filter | Options |
|---|---|
| Search | Free text by Name |
| Status | Available, Failed, Stopped, Creating, Stopping, Starting, Restarting, Terminating |
| Resource | GPU, CPU Only |
| Lock status | Locked, Unlocked |
Step 3. Workload Table
Default sort is newest first, 10 rows per page.
Table Columns
| Column | Description |
|---|---|
| Status | Pod execution state |
| Pod Name | Click to go to detail page |
| Label | ShortID derived from workload UUID |
| Lock Status | Lock state |
| Max timeout | Running timeout set at creation |
| Utilization | Real-time computing resource usage |
| Compute type | GPU name or "CPU Only" |
| Cost | Amount based on GPU/CPU selection |
| Action | Action dropdown |
Step 4. Pod Detail
Click the Pod Name or select Action → Detail to enter the detail page.
Monitoring Tab
| Item | Description |
|---|---|
| Disk Usage | used / total (%) |
| Volume Usage | Mounted volume used / total (%) |
| Uptime | Pod run time |
| Processes | Number of running processes |
| CPU Load | Usage (%) — half-donut chart |
| Memory | Usage (%) — half-donut chart |
Threshold colors: 0–69% green, 70–94% orange, 95%+ red
Log / Terminal Tab
Access Pod logs and a terminal from the detail page.
Step 5. Connect
Open the external access drawer via Action → Connect. Only active when the Pod is Running.
| Item | Description |
|---|---|
| Access URL | Cluster-specific domain via Control Plane |
| Open Service | Navigate to the service page |
| Set Up SSH Connection | Create an SSH session via Bastion host (5–10 seconds) |