flash-undeploy.md

flash undeploy

Manage and delete Runpod serverless endpoints deployed via Flash.

Overview

The flash undeploy command helps you clean up Serverless endpoints that Flash has created when you deployed Endpoint functions using flash dev or flash deploy. It manages endpoints recorded in .flash/resources.pkl and ensures both the cloud resources and local tracking state stay in sync.

When To Use This Command

• Cleaning up individual endpoints you no longer need
• Removing endpoints after local development/testing

flash undeploy vs flash env delete

Command	Scope	When to use
flash undeploy	Individual endpoints from local tracking	Granular cleanup, development endpoints
flash env delete	Entire environment + all its resources	Production cleanup, full teardown

For production deployments, use flash env delete to remove the entire environment and all associated resources automatically.

How Endpoint Tracking Works

Flash tracks deployed endpoints in .flash/resources.pkl. Endpoints get added to this file when you:

• Run flash dev --auto-provision (local development)
• Run flash deploy (production deployment)

Synopsis

flash undeploy [NAME|list] [OPTIONS]

Usage Modes

List Endpoints

Display all tracked endpoints with their current status:

flash undeploy list

Output includes:

• Name: Endpoint name
• Endpoint ID: Runpod endpoint identifier
• Status: Current health status (Active/Inactive/Unknown)
• Type: Resource type (Live Serverless, Cpu Live Serverless, etc.)
• Resource ID: Internal tracking identifier

Status Indicators:

• 🟢 Active: Endpoint is running and responding to health checks
• 🔴 Inactive: Endpoint tracking exists but health check fails (likely deleted externally)
• ❓ Unknown: Exception occurred during health check

Undeploy by Name

Delete a specific endpoint:

flash undeploy my-api

Behavior:

1. Searches for endpoints matching the name
2. Shows endpoint details
3. Prompts for confirmation
4. Deletes endpoint from Runpod
5. Removes from local tracking

Undeploy All

Delete all tracked endpoints with double confirmation:

flash undeploy --all

Safety features:

1. Shows total count of endpoints
2. First confirmation: Yes/No prompt
3. Second confirmation: Type "DELETE ALL" exactly
4. Deletes all endpoints from Runpod
5. Removes all from tracking

Interactive Selection

Select endpoints to undeploy using checkboxes:

flash undeploy --interactive

Behavior:

1. Displays interactive list with checkboxes
2. Use arrow keys to navigate
3. Use space bar to select/deselect
4. Press Enter to confirm selection
5. Prompts for final confirmation
6. Deletes selected endpoints

Cleanup Stale Tracking

Remove inactive endpoints from tracking without API deletion:

flash undeploy --cleanup-stale

Use case: When endpoints are deleted via Runpod UI or API (not through Flash), the tracking file becomes stale. This command identifies and removes those orphaned entries.

Behavior:

1. Checks health status of all tracked endpoints
2. Identifies inactive endpoints
3. Lists inactive endpoints for review
4. Prompts for confirmation
5. Removes only from local tracking (endpoints already deleted)

Options

• --all: Undeploy all endpoints (requires double confirmation)
• --interactive, -i: Interactive checkbox selection
• --cleanup-stale: Remove inactive endpoints from tracking

Examples

Basic Workflow

# Check what's deployed
flash undeploy list

# Remove a specific endpoint
flash undeploy my-api

# Clean up stale tracking
flash undeploy --cleanup-stale

Bulk Operations

# Undeploy all endpoints
flash undeploy --all

# Interactive selection for partial cleanup
flash undeploy --interactive

Managing External Deletions

If you delete endpoints via Runpod UI:

# Check status - will show as "Inactive"
flash undeploy list

# Remove stale tracking entries
flash undeploy --cleanup-stale

Understanding Status Checks

The Status column performs a health check API call for each endpoint. This:

• Adds latency (1 API call per endpoint)
• Provides accurate current state
• Identifies endpoints deleted externally

Why it's valuable:

• Catches endpoints deleted via Runpod UI
• Identifies unhealthy endpoints
• Prevents stale tracking file issues

Safety Features

Confirmations

• All deletion operations require explicit confirmation
• --all requires double confirmation with exact text match
• Interactive mode shows count before deletion

Warnings

• Clear indication that operations cannot be undone
• Displays endpoint details before deletion
• Shows success/failure counts after operations

Error Handling

• Continues processing remaining endpoints if one fails
• Shows detailed error messages per endpoint
• Tracks success/failure counts in summary

Tracking File

Endpoints are tracked in .flash/resources.pkl.

Important:

• This file is in .gitignore (never commit)
• Contains local deployment state
• Use flash undeploy --cleanup-stale to maintain accuracy
• make clean no longer deletes this file (use flash undeploy instead)

Integration with Endpoint

When you use the Endpoint class:

from runpod_flash import Endpoint, GpuGroup

@Endpoint(name="my-worker", gpu=GpuGroup.ANY)
def my_function(data):
    return {"result": data}

Flash automatically:

1. Deploys endpoint to Runpod
2. Tracks in .flash/resources.pkl
3. Reuses endpoint on subsequent calls

To clean up:

flash undeploy my-worker

Troubleshooting

Endpoint shows as "Inactive"

Cause: Endpoint was deleted via Runpod UI/API

Solution:

flash undeploy --cleanup-stale

Can't find endpoint by name

Symptoms: "No endpoint found with name 'my-api'"

Solution: Check exact name with:

flash undeploy list

Undeploy fails with API error

Cause: Network issues or invalid API key

Solution:

1. Check RUNPOD_API_KEY in .env
2. Verify network connectivity
3. Check endpoint still exists on Runpod

Related Commands

• flash init - Initialize new project
• flash dev - Start development server
• flash build - Build deployment packages
• flash deploy - Deploy to Runpod

See Also

• Flash CLI Overview
• Runpod Serverless Documentation
• Flash Documentation