epycloud status¶
Monitor active workflows and Cloud Batch jobs in real-time.
Usage¶
Description¶
Provides a real-time overview of active workflows and Cloud Batch jobs. The TASKS column shows completed work, while SLOTS shows occupied capacity against current demand. A separate provisioning warning identifies jobs that remain underfilled. Supports watch mode for continuous monitoring.
Options¶
| Option | Type | Description | Default |
|---|---|---|---|
--exp-id ID | Optional | Filter by experiment ID | All experiments |
--watch, -w | Flag | Watch mode - auto-refresh at interval | Disabled |
--recent [TIME], -r | Optional | Show recently completed items | 1 hour when present |
--interval N | Optional | Refresh interval in seconds (with --watch) | 10 |
--stall-threshold MINUTES | Optional | Warn after provisioning makes no progress for this long | 15 |
Examples¶
# Show status of all active workflows and jobs
epycloud status
# Show status for specific experiment
epycloud status --exp-id flu-2024
# Watch mode with default 10-second refresh
epycloud status --watch
# Watch with custom 5-second interval
epycloud status --watch --interval 5
# Use a 25-minute provisioning grace period
epycloud status --watch --stall-threshold 25
In one-shot mode, a never-started job can be timed from its creation time. A partially filled job is reported as underfilled with an unknown duration because one snapshot cannot establish when progress stopped. Watch mode retains count history and reports a stall only after neither completed tasks nor occupied slots have changed for the threshold. Provisioning warnings do not change the successful exit code.
Exit Codes¶
| Code | Description |
|---|---|
0 | Success |
1 | Error (API failure, invalid options) |
Related Commands¶
epycloud workflow list- List workflow executionsepycloud logs- View pipeline logsepycloud workflow describe- Detailed workflow info
See Also¶
- Monitoring Guide - Monitoring strategies and troubleshooting
- Google Cloud Batch