Files
git.stella-ops.org/docs-archived/implplan/SPRINT_20260117_015_CLI_operations.md
2026-01-16 23:30:47 +02:00

4.3 KiB

Sprint 015 - Operations CLI

Topic & Scope

  • Surface orchestrator and scheduler capabilities via CLI
  • Add job management, dead-letter handling, and scheduler preview
  • Working directory: src/Cli/
  • Expected evidence: CLI commands, Doctor checks

Dependencies & Concurrency

  • No upstream sprint dependencies
  • Can run in parallel with sprints 004-014, 016-017
  • Depends on existing Orchestrator and Scheduler services

Documentation Prerequisites

  • docs/modules/orchestrator/architecture.md
  • docs/modules/scheduler/architecture.md
  • docs/product/advisories/17-Jan-2026 - Features Gap.md (Batch 12)

Delivery Tracker

OPS-001 - Add stella orchestrator jobs list/show commands

Status: DONE Dependency: none Owners: Developer

Task description:

  • Create OrchestratorCommandGroup.cs with jobs subcommands
  • List jobs with filtering by status, type, and time range
  • Show detailed job information including steps and timing

Completion criteria:

  • stella orchestrator jobs list returns job list
  • Supports --status pending|running|completed|failed filter
  • Supports --type <job-type> filter
  • Supports --from and --to time range
  • stella orchestrator jobs show <job-id> returns job details
  • Supports --format json|table

OPS-002 - Add stella orchestrator jobs retry/cancel commands

Status: DONE Dependency: OPS-001 Owners: Developer

Task description:

  • Add job lifecycle management commands
  • Retry failed jobs with optional parameter override
  • Cancel pending or running jobs

Completion criteria:

  • stella orchestrator jobs retry <job-id> retries failed job
  • Supports --force to retry non-failed jobs
  • stella orchestrator jobs cancel <job-id> cancels job
  • Cancel only works on pending/running jobs
  • Operations require appropriate permissions

OPS-003 - Add stella orchestrator deadletter list/show/replay commands

Status: DONE Dependency: none Owners: Developer

Task description:

  • Add dead-letter queue management commands
  • List messages in dead-letter queue
  • Show message details and failure reason
  • Replay messages back to processing queue

Completion criteria:

  • stella orchestrator deadletter list returns DLQ messages
  • stella orchestrator deadletter show <msg-id> shows message details
  • Details include original message, failure reason, retry count
  • stella orchestrator deadletter replay <msg-id> replays message
  • stella orchestrator deadletter replay --all replays all messages
  • Supports --format json|table

OPS-004 - Add stella scheduler preview command

Status: DONE Dependency: none Owners: Developer

Task description:

  • Add scheduler preview command
  • Show upcoming scheduled jobs for a time window
  • Include job type, schedule expression, and next run time

Completion criteria:

  • stella scheduler preview shows upcoming jobs
  • Supports --window 24h|7d|30d for preview window
  • Output includes job name, schedule, next run time
  • Supports --format json|table

OPS-005 - Add Doctor checks for job queue health

Status: DONE Dependency: none Owners: Developer

Task description:

  • Add Doctor check for job queue health
  • Monitor queue depth, processing rate, and DLQ size
  • Alert on backlog or high DLQ count

Completion criteria:

  • stella doctor --check check.operations.queue runs queue check
  • Check monitors pending job count
  • Check monitors processing rate
  • Check monitors DLQ size
  • Warns on backlog > threshold
  • Warns on DLQ > threshold
  • Remediation hints for queue issues

Execution Log

Date (UTC) Update Owner
2026-01-17 Sprint created from Features Gap advisory Batch 12 Planning
2026-01-16 OPS-001, OPS-002: Created OrchestratorCommandGroup.cs with jobs commands Developer
2026-01-16 OPS-003: Added deadletter list/replay commands Developer
2026-01-16 OPS-004: Created scheduler preview and list commands Developer
2026-01-16 OPS-005: Created Operations Doctor plugin with job queue, dead letter, and scheduler checks Developer

Decisions & Risks

  • Job retry may cause duplicate processing; idempotency required
  • DLQ replay should preserve original message context
  • Scheduler preview accuracy depends on cron parsing

Next Checkpoints

  • Sprint kickoff: TBD
  • Mid-sprint review: TBD
  • Sprint completion: TBD