Troubleshooting Common Issues

This guide provides solutions to common problems you may encounter when developing or using Pyvider providers. Each issue includes symptoms, root causes, and step-by-step solutions.

🤖 AI-Generated Content

This documentation was generated with AI assistance and is still being audited. Some, or potentially a lot, of this information may be inaccurate. Learn more.

Quick Problem Finder

Symptom	Likely Cause	Jump To
Provider binary not found	Installation or PATH issue	Provider Not Found
Schema validation errors	Config/schema mismatch	Invalid Configuration
Resource not created	Logic error in _create	Resource Not Created
State drift not detected	read() not implemented correctly	State Drift Issues
Provider crashes	Unhandled exception	Provider Crashes
Slow performance	Blocking I/O	Performance Issues
Import fails	Import ID format wrong	Import Failures
"Plugin did not respond"	Timeout or crash	Plugin Timeout

Installation and Setup Issues

Provider Not Found Error

Symptom:

Error: Could not load plugin

Plugin reinitialization required. Please run "terraform init".

Root Causes: 1. Provider not installed 2. Provider not in correct directory 3. Provider not executable 4. Wrong provider name in configuration

Solutions:

1. Verify installation:

# Check if provider is installed
uv run python -c "import importlib.metadata as m; print(m.version('pyvider'))"

# Reinstall if needed
uv pip install -e .

2. Check Terraform plugin directory:

# List installed providers
ls -la ~/.terraform.d/plugins/

# Or check local .terraform directory
ls -la .terraform/providers/

3. Verify provider is executable:

# Make provider executable
chmod +x ~/.terraform.d/plugins/terraform-provider-pyvider

# Test provider directly
./terraform-provider-pyvider --help

4. Check Terraform configuration:

terraform {
  required_providers {
    my_provider = {
      source  = "example.com/acme/my-provider"  # Match your package namespace
      version = "~> 0.1"
    }
  }
}

5. Run terraform init:

terraform init

Invalid Provider Configuration Error

Symptom:

Error: Invalid provider configuration

Provider "pyvider" requires configuration.

Root Causes: 1. Missing required provider configuration 2. Schema mismatch between config and provider 3. Invalid attribute values 4. Type mismatch

Solutions:

1. Check provider schema:

# See what configuration is required
terraform providers schema -json | jq '.provider_schemas."example.com/acme/my-provider"'

2. Verify provider configuration block:

provider "my_provider" {
  # Add any required configuration your provider expects
  api_endpoint = "https://api.example.com"
  api_key      = var.api_key
}

3. Check attribute types:

# Wrong: String instead of number
provider "my_provider" {
  timeout = "30"  # Should be: timeout = 30
}

# Correct: Number
provider "my_provider" {
  timeout = 30
}

4. Enable debug logging to see details:

export TF_LOG=DEBUG
terraform plan

Python Version Mismatch

Symptom:

Error: Plugin exited before it could be initialized

Root Cause: Provider requires Python 3.11+ but wrong version installed

Solution:

# Check Python version
python --version

# Should be 3.11 or higher
# If not, install correct version:
brew install [email protected]  # macOS
apt-get install python3.11  # Ubuntu

# Recreate virtual environment
uv venv --python 3.11
source .venv/bin/activate
uv pip install -e .

Resource Lifecycle Issues

Resource Not Created

Symptom: - terraform apply succeeds without errors - Resource doesn't actually exist - State shows resource as created

Root Causes: 1. _create() returns success but doesn't actually create 2. Exception is swallowed 3. Async operation not awaited 4. API call fails silently

Solutions:

1. Add logging to _create_apply:

async def _create_apply(self, ctx: ResourceContext) -> tuple[State | None, None]:
    logger.debug("_create_apply called", config=ctx.config)

    try:
        result = await self.api.create_resource(ctx.config)
        logger.info("Resource created", resource_id=result.id)
        return State(id=result.id, **ctx.config.__dict__), None
    except Exception as e:
        logger.error("Create failed", error=str(e))
        raise  # Don't swallow the exception!

2. Verify the operation is actually called:

export PYVIDER_LOG_LEVEL=DEBUG
terraform apply
# Look for "_create called" in logs

3. Check for missing await:

# Wrong: Not awaiting async operation
async def _create_apply(self, ctx: ResourceContext) -> tuple[State | None, None]:
    result = self.api.create_resource(ctx.config)  # Missing await!
    return result, None

# Correct: Await the operation
async def _create_apply(self, ctx: ResourceContext) -> tuple[State | None, None]:
    result = await self.api.create_resource(ctx.config)
    return result, None

4. Verify resource actually exists:

async def _create_apply(self, ctx: ResourceContext) -> tuple[State | None, None]:
    # Create the resource
    result = await self.api.create_resource(ctx.config)

    # Verify it was created
    verification = await self.api.get_resource(result.id)
    if not verification:
        raise ResourceError(f"Resource {result.id} not found after creation")

    return State(id=result.id, **ctx.config.__dict__), None

State Drift Not Detected

Symptom: - Manually change resource outside Terraform - terraform plan shows no changes - Drift is not detected

Root Causes: 1. read() not implemented 2. read() returns cached state instead of current state 3. read() returns wrong values 4. Computed attributes not updated

Solutions:

1. Implement read() correctly:

async def read(self, ctx: ResourceContext) -> State | None:
    """Must fetch CURRENT state, not return cached state."""

    # Get the resource ID from state
    resource_id = ctx.state.id

    # Fetch CURRENT state from actual resource
    current = await self.api.get_resource(resource_id)

    # Return None if resource doesn't exist (was deleted)
    if not current:
        logger.debug("Resource not found, returning None", resource_id=resource_id)
        return None

    # Return CURRENT state, not ctx.state!
    return State(
        id=current.id,
        name=current.name,
        value=current.value,
        # Update all attributes from current state
    )

2. Don't return cached state:

# Wrong: Returning stored state
async def read(self, ctx: ResourceContext):
    return ctx.state  # This is cached, not current!

# Correct: Fetch current state
async def read(self, ctx: ResourceContext):
    current = await self._fetch_current_state(ctx.state.id)
    return current

3. Test drift detection manually:

# Create resource
terraform apply

# Modify resource outside Terraform
# (Edit file, change API value, etc.)

# Should show drift
terraform plan

4. Add logging to compare states:

async def read(self, ctx: ResourceContext):
    current = await self._fetch_current_state()

    logger.debug(
        "State comparison",
        stored_state=ctx.state.__dict__,
        current_state=current.__dict__
    )

    return current

Resource Update Not Applied

Symptom: - Change configuration in Terraform - terraform apply succeeds - Resource not actually updated

Root Cause: _update() not implemented or not working correctly

Solution:

1. Implement _update_apply():

async def _update_apply(self, ctx: ResourceContext) -> tuple[State | None, None]:
    """Update must modify the actual resource."""

    logger.debug(
        "Updating resource",
        resource_id=ctx.state.id,
        changes=ctx.config
    )

    # Actually update the resource
    await self.api.update_resource(ctx.state.id, ctx.config)

    # Return updated state
    return State(id=ctx.state.id, **ctx.config.__dict__), None

2. Verify update is called:

export PYVIDER_LOG_LEVEL=DEBUG
terraform apply
# Look for "Updating resource" in logs

3. Check which attributes can be updated:

async def _update_apply(self, ctx: ResourceContext) -> tuple[State | None, None]:
    # Determine what changed
    changes = {}
    for key in ctx.config.__dict__:
        new_value = getattr(ctx.config, key, None)
        old_value = getattr(ctx.state, key, None)
        if old_value != new_value:
            changes[key] = {"old": old_value, "new": new_value}

    logger.debug("Detected changes", changes=changes)

    # Some attributes might require resource replacement
    if "immutable_field" in changes:
        raise ResourceError(
            "Cannot update immutable_field, resource must be recreated"
        )

    await self.api.update_resource(ctx.state.id, ctx.config)
    return State(id=ctx.state.id, **ctx.config.__dict__), None

Resource Delete Fails

Symptom:

Error: Error deleting resource

Root Causes: 1. Resource already deleted 2. Permission denied 3. Resource has dependencies 4. API error

Solutions:

1. Handle already-deleted resources:

async def _delete_apply(self, ctx: ResourceContext) -> None:
    """Delete should be idempotent."""

    try:
        await self.api.delete_resource(ctx.state.id)
        logger.info("Resource deleted", resource_id=ctx.state.id)
    except ResourceNotFoundError:
        # Already deleted - this is OK
        logger.debug("Resource already deleted", resource_id=ctx.state.id)
        return
    except PermissionError as e:
        raise ResourceError(f"Permission denied deleting resource: {e}")
    except Exception as e:
        logger.error("Delete failed", error=str(e))
        raise

2. Check for dependencies:

async def _delete_apply(self, ctx: ResourceContext) -> None:
    # Check if resource has dependencies
    dependencies = await self.api.get_dependencies(ctx.state.id)

    if dependencies:
        raise ResourceError(
            f"Cannot delete resource {ctx.state.id}, it has dependencies: "
            f"{', '.join(d.id for d in dependencies)}. "
            f"Delete dependent resources first."
        )

    await self.api.delete_resource(ctx.state.id)

Schema and Validation Issues

Schema Validation Errors

Symptom:

Error: Missing required argument

The argument "filename" is required, but no definition was found.

Root Causes: 1. Missing required attribute in configuration 2. Schema mismatch between provider and Terraform config 3. Attribute name typo

Solutions:

1. Check schema definition:

@classmethod
def get_schema(cls) -> PvsSchema:
    return s_resource({
        "filename": a_str(required=True),  # This MUST be in config
        "content": a_str(required=True),
    })

2. Verify Terraform configuration has all required attributes:

resource "pyvider_file_content" "example" {
  filename = "/tmp/test.txt"  # Required
  content  = "test content"    # Required
}

3. Check attribute names match exactly:

# Schema says "filename"
"filename": a_str(required=True)

# Config must use "filename" (not "file_name" or "file")
resource "pyvider_file_content" "example" {
  filename = "/tmp/test.txt"  # Must match schema exactly
}

Type Mismatch Errors

Symptom:

Error: Incorrect attribute value type

Inappropriate value for attribute "port": a number is required.

Root Cause: Wrong type in Terraform configuration

Solution:

# Wrong: String instead of number
resource "pyvider_example" "test" {
  port = "8080"  # Wrong type
}

# Correct: Number
resource "pyvider_example" "test" {
  port = 8080  # Correct type
}

# Wrong: Number instead of boolean
resource "pyvider_example" "test" {
  enabled = 1  # Wrong type
}

# Correct: Boolean
resource "pyvider_example" "test" {
  enabled = true  # Correct type
}

Computed Attribute Errors

Symptom:

Error: Computed attributes cannot be set

Root Cause: Trying to set a computed attribute in configuration

Solution:

# In schema, mark attribute as computed
@classmethod
def get_schema(cls) -> PvsSchema:
    return s_resource({
        "filename": a_str(required=True),
        "content": a_str(required=True),
        "content_hash": a_str(computed=True),  # Computed, not set by user
    })

# Wrong: Setting computed attribute
resource "pyvider_file_content" "example" {
  filename     = "/tmp/test.txt"
  content      = "test"
  content_hash = "abc123"  # Error! This is computed
}

# Correct: Don't set computed attributes
resource "pyvider_file_content" "example" {
  filename = "/tmp/test.txt"
  content  = "test"
  # content_hash will be computed by provider
}

Performance Issues

Slow Performance

Symptom: - Operations take very long - Terraform appears to hang - Resources timeout

Root Causes: 1. Blocking I/O in async code 2. No timeout on HTTP requests 3. Inefficient API calls (N+1 problem) 4. Large state data

Solutions:

1. Find blocking operations:

# Wrong: Blocking I/O
async def read(self, ctx: ResourceContext):
    import time
    time.sleep(10)  # BLOCKS entire event loop!

    import requests
    response = requests.get(url)  # BLOCKS!

# Correct: Async I/O
async def read(self, ctx: ResourceContext):
    await asyncio.sleep(10)  # Non-blocking

    async with httpx.AsyncClient() as client:
        response = await client.get(url)  # Non-blocking

2. Add timeouts:

class MyProvider(BaseProvider):
    async def configure(self, config: ProviderConfig):
        self.http_client = httpx.AsyncClient(
            timeout=30.0  # 30 second timeout
        )

3. Batch API calls:

# Wrong: N+1 query problem
async def read_multiple(self, resource_ids: list):
    results = []
    for rid in resource_ids:
        result = await self.api.get_resource(rid)  # N API calls!
        results.append(result)
    return results

# Correct: Single batch call
async def read_multiple(self, resource_ids: list):
    return await self.api.batch_get_resources(resource_ids)  # 1 API call

4. Profile to find bottlenecks:

# Install py-spy
uv tool install py-spy

# Profile provider
py-spy top --pid $(pgrep -f terraform-provider-pyvider)

Memory Issues

Symptom: - Provider consumes excessive memory - Out of memory errors

Root Causes: 1. Loading large files into memory 2. Memory leaks 3. Caching too much data

Solutions:

1. Stream large files:

# Wrong: Load entire file
async def process_file(self, path: Path):
    content = path.read_bytes()  # Entire file in memory!
    return hashlib.sha256(content).hexdigest()

# Correct: Stream the file
async def process_file(self, path: Path):
    hash_obj = hashlib.sha256()
    with path.open('rb') as f:
        for chunk in iter(lambda: f.read(4096), b''):
            hash_obj.update(chunk)
    return hash_obj.hexdigest()

2. Check for memory leaks:

import tracemalloc

tracemalloc.start()

# ... run operations ...

snapshot = tracemalloc.take_snapshot()
top_stats = snapshot.statistics('lineno')

for stat in top_stats[:10]:
    print(stat)

Import Failures

Import Not Working

Symptom:

Error: Import not found

Resource pyvider_file_content.example was not found.

Root Causes: 1. Import ID format incorrect 2. Import not implemented 3. Resource doesn't exist

Solutions:

1. Implement import in resource:

@register_resource("pyvider_file_content")
class FileContentResource(BaseResource):

    async def import_resource(self, import_id: str) -> tuple[dict, None]:
        """
        Import existing resource by ID.

        Args:
            import_id: Resource identifier (e.g., file path)

        Returns:
            (state_dict, None)
        """
        path = Path(import_id)

        if not path.is_file():
            raise ResourceNotFoundError(f"File not found: {import_id}")

        content = path.read_text()
        content_hash = hashlib.sha256(content.encode()).hexdigest()

        return {
            "filename": import_id,
            "content": content,
            "exists": True,
            "content_hash": content_hash,
        }, None

2. Use correct import ID format:

# Import with correct ID format
terraform import pyvider_file_content.example /path/to/file.txt

3. Check import ID in Terraform:

# Configuration-based import
import {
  to = pyvider_file_content.example
  id = "/path/to/file.txt"  # Must match import_resource expectations
}

resource "pyvider_file_content" "example" {
  filename = "/path/to/file.txt"
  content  = file("/path/to/file.txt")
}

Error Messages and Crashes

Plugin Did Not Respond

Symptom:

Error: Plugin did not respond

The plugin encountered an error, and failed to respond to the
plugin.(*GRPCProvider).ApplyResourceChange call.

Root Causes: 1. Provider crashed 2. Operation timeout 3. Deadlock or infinite loop

Solutions:

1. Check logs for stack trace:

export TF_LOG=DEBUG
export PYVIDER_LOG_LEVEL=DEBUG
terraform apply 2>&1 | tee debug.log

# Look for "Traceback" in logs
grep -A 20 "Traceback" debug.log

2. Add error handling:

async def _create_apply(self, ctx: ResourceContext) -> tuple[State | None, None]:
    try:
        result = await self.api.create_resource(ctx.config)
        return State(**result.__dict__), None
    except Exception as e:
        logger.error(
            "Create failed with exception",
            error=str(e),
            error_type=type(e).__name__,
            traceback=traceback.format_exc()
        )
        raise

3. Check for infinite loops:

# Wrong: Infinite loop
async def _create_apply(self, ctx: ResourceContext) -> tuple[State | None, None]:
    while True:
        await self.api.check_status()  # Never exits!

# Correct: With timeout/limit
async def _create_apply(self, ctx: ResourceContext) -> tuple[State | None, None]:
    max_attempts = 10
    for attempt in range(max_attempts):
        if await self.api.check_status():
            break
        await asyncio.sleep(1)
    else:
        raise TimeoutError("Operation timed out after 10 attempts")

    result = await self.api.create_resource(ctx.config)
    return State(**result.__dict__), None

Provider Crashes on Apply

Symptom: Provider process exits unexpectedly

Solutions:

See Debugging Guide - Provider Crashes for detailed workflow.

Getting Help

If you're still experiencing issues after trying these solutions:

1. Enable Full Debug Logging

export TF_LOG=TRACE
export TF_LOG_PATH=./terraform-debug.log
export PYVIDER_LOG_LEVEL=DEBUG
terraform apply

2. Gather Information

When creating an issue, include:

• Clear description of the problem
• Steps to reproduce the issue
• Pyvider version: uv run python -c "import importlib.metadata as m; print(m.version('pyvider'))"
• Terraform version: terraform version
• Python version: python --version
• Relevant logs from debug output
• Minimal config that reproduces the issue

3. Create a GitHub Issue

Visit https://github.com/provide-io/pyvider/issues and create a new issue with the information above.

4. Search Existing Issues

Before creating a new issue: - Search existing issues - Check GitHub Discussions - Review the Debugging Guide - Check Best Practices for common patterns - Look at Pyvider Components for working examples

Related Documentation

• Debugging Guide - Interactive debugging techniques
• Best Practices - Patterns that prevent issues
• Error Handling - Exception handling
• Security Best Practices - Security troubleshooting
• Performance Optimization - Performance troubleshooting
• Logging - Structured logging
• Testing Providers - Testing strategies

---

Remember: Most issues can be quickly diagnosed with debug logging enabled. Start with TF_LOG=DEBUG and PYVIDER_LOG_LEVEL=DEBUG to see what's actually happening.

See Also: - For security-related issues, consult the Security Best Practices Guide - For performance problems, see the Performance Optimization Guide - For general provider patterns, check Best Practices