Troubleshooting guide¶

Common issues, error messages, and solutions for GlassAlpha. This guide helps diagnose and resolve problems quickly.

Quick diagnostics¶

If you're experiencing issues, start with these diagnostic commands:

# Verify installation (if package is installed)
glassalpha --version

# Alternative: Use module invocation (development/uninstalled)
PYTHONPATH=src python3 -m glassalpha --version

# Check component availability
glassalpha list
# Alternative: PYTHONPATH=src python3 -m glassalpha list

# Validate configuration
glassalpha validate --config german_credit_simple.yaml
# Alternative: PYTHONPATH=src python3 -m glassalpha validate --config german_credit_simple.yaml

# Test with minimal configuration
glassalpha audit --config german_credit_simple.yaml --output test.pdf --dry-run
# Alternative: PYTHONPATH=src python3 -m glassalpha audit --config german_credit_simple.yaml --output test.pdf --dry-run

CLI command issues¶

`glassalpha: command not found`¶

This indicates GlassAlpha isn't properly installed or the CLI entry point isn't available.

Solutions:

Install the package (recommended for users):

cd glassalpha
pip install -e .

# Verify installation
glassalpha --version

Use module invocation (development/troubleshooting):

cd glassalpha
PYTHONPATH=src python3 -m glassalpha --version
PYTHONPATH=src python3 -m glassalpha audit --config german_credit_simple.yaml --output test.pdf --dry-run

Check your environment:

# Check if package is installed
pip list | grep glassalpha

# Check Python path
python3 -c "import sys; print('\n'.join(sys.path))"

Installation issues¶

Python version compatibility¶

Error:

ERROR: Package 'glassalpha' requires a different Python: 3.10.0 not in '>=3.11'

Solution:

# Check Python version
python --version

# Upgrade Python (macOS with Homebrew)
brew install python@3.11
python3.11 -m pip install -e .

# Upgrade Python (Linux)
sudo apt update && sudo apt install python3.11

# Create virtual environment with correct Python
python3.11 -m venv venv
source venv/bin/activate
pip install -e .

XGBoost installation issues (macOS)¶

Error:

RuntimeError: libomp not found. Install with: brew install libomp

Solution:

# Install OpenMP library
brew install libomp

# Reinstall XGBoost
pip uninstall xgboost
pip install xgboost

# Verify installation
python -c "import xgboost; print('XGBoost version:', xgboost.__version__)"

Dependency conflicts¶

Error:

ERROR: pip's dependency resolver does not currently take into account all the packages that are installed

Solution:

# Create clean virtual environment
python -m venv clean_env
source clean_env/bin/activate  # On Windows: clean_env\Scripts\activate

# Upgrade pip
pip install --upgrade pip

# Install GlassAlpha
pip install -e .

# If conflicts persist, install dependencies individually
pip install pandas numpy scikit-learn
pip install xgboost lightgbm
pip install shap matplotlib seaborn
pip install -e . --no-deps

Missing system dependencies¶

Error (Linux):

ImportError: libgomp.so.1: cannot open shared object file

Solution (Ubuntu/Debian):

sudo apt update
sudo apt install libgomp1 build-essential

Solution (CentOS/RHEL):

sudo yum install libgomp gcc gcc-c++

Configuration errors¶

Missing required fields¶

Error:

ValidationError: 1 validation error for AuditConfig
data.target_column
  field required (type=value_error.missing)

Solution:

# Add missing required fields
data:
  path: data/dataset.csv
  target_column: outcome # This was missing

model:
  type: xgboost # Ensure model type is specified

Invalid file paths¶

Error:

FileNotFoundError: Configuration file 'nonexistent.yaml' not found

Solutions:

# Use absolute paths
glassalpha audit --config /full/path/to/config.yaml --output report.pdf

# Config files are packaged with GlassAlpha - no need for configs/ directory
# Just reference by name:
glassalpha audit --config german_credit_simple.yaml --output audit.pdf

Model type not found¶

Error:

Warning: Model type 'unknown_model' not supported

Solution:

# Check available components (models, explainers, metrics)
glassalpha list

# Use correct model type
# Valid options: xgboost, lightgbm, logistic_regression, sklearn_generic

model:
  type: xgboost # Use registered model type

Schema validation errors¶

Error:

ValidationError: 2 validation errors for TabularDataSchema
features
  ensure this value has at least 1 characters
target
  field required

Solution:

data:
  path: data/dataset.csv
  target_column: target # Must specify target
  feature_columns: # Either specify features explicitly
    - feature1
    - feature2
  # OR let GlassAlpha auto-detect (remove feature_columns)

Model and dependency issues¶

Model not available errors¶

Error messages:

Model 'xgboost' not available. Falling back to 'logistic_regression'.
To enable 'xgboost', run: pip install 'glassalpha[xgboost]'

Or:

Missing optional dependency 'xgboost' for plugin 'xgboost'.
Try: pip install 'glassalpha[xgboost]'

Cause: GlassAlpha uses optional dependencies to keep the core installation lightweight. XGBoost and LightGBM are not installed by default.

Solutions:

Use the baseline model (recommended for getting started):

# In your config.yaml
model:
  type: logistic_regression # Always available
  params:
    random_state: 42

Install the features you need:

# For SHAP + tree models (includes XGBoost and LightGBM)
pip install 'glassalpha[explain]'

# For all features
pip install 'glassalpha[all]'

Check what's currently available:

glassalpha doctor

Disable fallbacks for strict requirements:

model:
  type: xgboost
  allow_fallback: false # Fail if XGBoost not available

Model loading failures¶

Error messages:

ModuleNotFoundError: No module named 'xgboost'

Cause: The model library isn't installed in your environment.

Solutions: Same as above - install the required optional dependency.

Fallback model not suitable¶

Error messages:

Model 'xgboost' not available. Falling back to 'logistic_regression'.

Cause: The requested model isn't available, and fallback is enabled.

Solutions:

Accept the fallback (LogisticRegression is a solid baseline)
Install the requested model using the provided installation command

Disable fallbacks if you need the specific model:

model:
  type: xgboost
  allow_fallback: false

Explainer errors (SHAP library required)¶

Error messages:

SHAP library required for KernelSHAP explainer

Or:

No explainer from ['treeshap'] is available. Missing dependencies: ['shap'].
Install with: pip install 'glassalpha[explain]'
OR use zero-dependency explainers: ['coefficients', 'permutation']

Cause: Your configuration requests a SHAP-based explainer (TreeSHAP or KernelSHAP) but the SHAP library isn't installed in your environment.

Solutions:

Use zero-dependency explainers (recommended for getting started):

# In your config.yaml
explainers:
  priority:
    - coefficients # Best for LogisticRegression (no dependencies)
    - permutation # Works with any model (no dependencies)

Install SHAP for tree model explanations:

# Install SHAP + XGBoost + LightGBM
pip install 'glassalpha[explain]'

# Verify installation
glassalpha doctor

Use the quickstart config which works without SHAP:

glassalpha audit --config quickstart.yaml --output audit.html

Which explainer should I use?

Model Type	Best Explainer	Requires SHAP?	Notes
LogisticRegression	`coefficients`	❌ No	Zero dependencies, fast
XGBoost/LightGBM	`treeshap`	✅ Yes	Most accurate for tree models
RandomForest	`treeshap`	✅ Yes	Handles tree ensembles well
Any model	`permutation`	❌ No	Universal fallback, model-agnostic

Example configs:

For LogisticRegression (works with base install):

model:
  type: logistic_regression
explainers:
  priority:
    - coefficients

For XGBoost (requires pip install 'glassalpha[explain]'):

model:
  type: xgboost
explainers:
  priority:
    - treeshap
    - permutation # Fallback if SHAP fails

Data issues¶

Data format mismatch in reasons/recourse commands¶

Error:

Error: Test data missing 15 columns.
Expected 22 columns: age_group, checking_account_status, ...
Actual 67 columns: age_group_middle_aged, age_group_middle_young, ...

Cause:

The reasons and recourse commands expect data in the same format the model was trained with (including preprocessing/one-hot encoding). If you provide the original CSV data, it won't match the model's expected format.

Solution:

Always use the preprocessed test data saved during audit generation:

# ✅ CORRECT - Use preprocessed test data
glassalpha reasons \
  --model models/german_credit.pkl \
  --data models/test_data.csv \
  --instance 0

# ❌ WRONG - Original data needs preprocessing
glassalpha reasons \
  --model models/german_credit.pkl \
  --data data/german_credit.csv \
  --instance 0

Why this happens:

During audit generation, GlassAlpha applies preprocessing (one-hot encoding categorical features)
The model is trained on this preprocessed data
The preprocessed test data is saved to models/test_data.csv
Reason codes/recourse need the same preprocessed format

Automatic detection (v0.2.0+):

GlassAlpha now automatically detects if data is already preprocessed and skips double preprocessing. You'll see:

✓ Data is already preprocessed (matches model's expected features)

If you see "Applying preprocessing to match model training..." but still get errors, use the test_data.csv file instead.

Dataset not found¶

Error:

FileNotFoundError: Dataset file 'data/missing.csv' not found

Solutions:

Check file paths (most common issue):

# Check file existence
ls -la data/missing.csv

# Check current directory
pwd

# Use absolute paths in config

data:
  path: /Users/yourname/data/dataset.csv # Absolute path
  # OR
  path: ~/data/dataset.csv # Home directory expansion

Common Mistake: Relative Paths

Relative paths are resolved from where you run the command, not where the config file is located. Use absolute paths or ~ for home directory.

Use built-in datasets for testing:

data:
  source: german_credit # Built-in dataset (auto-downloaded)
  fetch: if_missing

Browse available datasets:

glassalpha datasets list

Need help with data preparation? See Using Custom Data for a complete tutorial.

Column not found¶

Error:

KeyError: Column 'target' not found in dataset

Causes:

Column name misspelled or incorrect case
CSV file has no headers
Column name has spaces or special characters

Solutions:

Inspect your data:

# Check column names
python -c "import pandas as pd; df = pd.read_csv('data/dataset.csv'); print('Columns:', df.columns.tolist()); print('Shape:', df.shape)"

# Check first few rows
python -c "import pandas as pd; print(pd.read_csv('data/dataset.csv').head())"

Update configuration with exact column name:

data:
  path: data/dataset.csv
  target_column: "actual_column_name" # Use quotes if spaces/special chars


  # Check for common issues:
  # - Extra spaces: "target " vs "target"
  # - Case sensitivity: "Target" vs "target"
  # - Special characters: "target_column" vs "target-column"

Fix CSV headers:

import pandas as pd

# Read CSV without headers
df = pd.read_csv('data/dataset.csv', header=None)

# Add column names
df.columns = ['feature1', 'feature2', 'target']

# Save with headers
df.to_csv('data/dataset_with_headers.csv', index=False)

Protected attributes not found¶

Error:

KeyError: Protected attribute 'gender' not found in dataset

Solutions:

Check attribute names:

# List all columns
python -c "import pandas as pd; print(pd.read_csv('data/dataset.csv').columns.tolist())"

Update config with correct names:

data:
  protected_attributes:
    - sex # Not 'gender'
    - age_group # Not 'age'

If attributes are missing from your data:

data:
  protected_attributes: [] # Empty list if no protected attributes available

# Note: Fairness metrics will be limited without protected attributes

For more on protected attributes: See Using Custom Data.

Data format issues¶

Error:

ParserError: Error tokenizing data. C error: Expected 5 fields, saw 7

Causes:

Inconsistent number of columns
Unescaped commas in text fields
Corrupted file

Solutions:

Check for delimiter issues:

import pandas as pd

# Try different delimiters
df = pd.read_csv('data/dataset.csv', delimiter='\t')  # Tab-separated
df = pd.read_csv('data/dataset.csv', delimiter=';')   # Semicolon-separated

Handle problematic CSV files:

# Read with error handling
df = pd.read_csv('data/dataset.csv',
                 on_bad_lines='skip',  # Skip bad lines
                 encoding='utf-8')      # Specify encoding

Convert to a supported format:

import pandas as pd

# Read problematic CSV
df = pd.read_csv('data/dataset.csv', on_bad_lines='skip')

# Save as Parquet (more robust)
df.to_parquet('data/dataset.parquet', index=False)

# Update config to use Parquet
data:
  path: data/dataset.parquet

Data type issues¶

Error:

TypeError: Object of type 'datetime64' is not JSON serializable

Causes:

Datetime columns not converted
Complex object types
Mixed data types in columns

Solutions:

Convert datetime columns:

import pandas as pd

df = pd.read_csv('data/dataset.csv')

# Option 1: Convert to Unix timestamp
df['date_column'] = pd.to_datetime(df['date_column']).astype('int64') // 10**9

# Option 2: Extract useful features
df['year'] = pd.to_datetime(df['date_column']).dt.year
df['month'] = pd.to_datetime(df['date_column']).dt.month
df['day_of_week'] = pd.to_datetime(df['date_column']).dt.dayofweek

# Drop original datetime column
df = df.drop('date_column', axis=1)

df.to_csv('data/processed_dataset.csv', index=False)

Handle categorical data:

# Check data types
print(df.dtypes)

# Convert object columns to category or numeric
df['category_column'] = df['category_column'].astype('category').cat.codes

Identify problematic columns:

# Find columns with mixed types
for col in df.columns:
    unique_types = df[col].apply(type).unique()
    if len(unique_types) > 1:
        print(f"Column '{col}' has mixed types: {unique_types}")

Missing values¶

Error:

ValueError: Input contains NaN, infinity or a value too large

Causes:

Missing data in dataset
Infinite values from calculations
Very large numbers causing overflow

Solutions:

Inspect missing data:

import pandas as pd

df = pd.read_csv('data/dataset.csv')

# Check missing values per column
print(df.isnull().sum())

# Check percentage missing
print(df.isnull().sum() / len(df) * 100)

# Identify rows with missing target
print(df[df['target'].isnull()])

Handle missing values:

# Option 1: Drop rows with missing target (recommended)
df = df.dropna(subset=['target'])

# Option 2: Drop columns with >30% missing
threshold = 0.3
df = df.loc[:, df.isnull().mean() < threshold]

# Option 3: Impute missing values
from sklearn.impute import SimpleImputer

# For numeric columns
numeric_cols = df.select_dtypes(include=['float64', 'int64']).columns
imputer = SimpleImputer(strategy='median')
df[numeric_cols] = imputer.fit_transform(df[numeric_cols])

# For categorical columns
categorical_cols = df.select_dtypes(include=['object']).columns
df[categorical_cols] = df[categorical_cols].fillna(df[categorical_cols].mode().iloc[0])

df.to_csv('data/cleaned_dataset.csv', index=False)

Check for infinite values:

import numpy as np

# Replace infinite values
df = df.replace([np.inf, -np.inf], np.nan)

# Then handle as missing values
df = df.dropna()

More data preparation guidance: See Using Custom Data - Data Cleaning.

Target column issues¶

Error:

ValueError: Target column must be binary (0/1) for classification

Causes:

Target has more than 2 unique values
Target is text instead of numeric
Target has missing values

Solutions:

Check target distribution:

import pandas as pd

df = pd.read_csv('data/dataset.csv')
print("Target values:", df['target'].unique())
print("Target counts:", df['target'].value_counts())
print("Target type:", df['target'].dtype)

Convert target to binary:

# If target is text (e.g., "Yes"/"No")
df['target'] = (df['target'] == 'Yes').astype(int)

# If target is numeric but not 0/1
df['target'] = (df['target'] > threshold).astype(int)

# If target is multi-class, convert to binary
# Example: predict if class is "high risk" vs others
df['target'] = (df['target'] == 'high_risk').astype(int)

df.to_csv('data/dataset_binary.csv', index=False)

Remove missing target values:

df = df.dropna(subset=['target'])
print(f"Rows after removing missing targets: {len(df)}")

Insufficient data¶

Error:

ValueError: Insufficient samples for training (found 50, need at least 100)

Causes:

Dataset too small
Too many missing values removed
Heavy class imbalance with very few positive samples

Solutions:

Check dataset size:

import pandas as pd

df = pd.read_csv('data/dataset.csv')
print(f"Total rows: {len(df)}")
print(f"Class distribution:\n{df['target'].value_counts()}")
print(f"Minimum class size: {df['target'].value_counts().min()}")

Use a larger dataset:
Browse built-in datasets
Combine multiple datasets if possible
Use German Credit (1,000 rows) for testing
If stuck with small data:

# Adjust model parameters for small datasets
model:
  type: logistic_regression # Better for small data than tree models
  params:
    max_iter: 1000
    C: 1.0

# Reduce explainer sample sizes
explainers:
  config:
    treeshap:
      max_samples: 50 # Reduce to match data size

Class imbalance¶

Warning:

Warning: Severe class imbalance detected (99.5% negative class)

Solutions:

Check imbalance ratio:

import pandas as pd

df = pd.read_csv('data/dataset.csv')
value_counts = df['target'].value_counts()
imbalance_ratio = value_counts.max() / value_counts.min()
print(f"Imbalance ratio: {imbalance_ratio:.1f}:1")
print(f"Class distribution:\n{value_counts}")

Adjust model for imbalance:

model:
  type: xgboost
  params:
    # Critical for imbalanced data
    scale_pos_weight: 99 # Ratio of negative to positive samples
    objective: binary:logistic

Use appropriate metrics:

metrics:
  performance:
    metrics:
      - precision # More important than accuracy for imbalanced data
      - recall
      - f1
      - pr_auc # Better than roc_auc for imbalanced data

Example configurations: See credit_card_fraud_template.yaml (packaged with GlassAlpha) for handling 99.8% class imbalance.

Runtime errors¶

Memory issues¶

Error:

MemoryError: Unable to allocate array with shape (10000, 10000)

Solutions:

# Reduce sample sizes for explainers
explainers:
  config:
    treeshap:
      max_samples: 100 # Reduce from default 1000
    kernelshap:
      n_samples: 50 # Reduce from default 500

# Enable low memory mode
performance:
  low_memory_mode: true
  n_jobs: 1 # Reduce parallelism

# Use smaller datasets for testing
head -n 1000 large_dataset.csv > small_dataset.csv

Model training failures¶

Error:

XGBoostError: Check failed: labels.Size() == num_row

Solution:

# Check data consistency
import pandas as pd
df = pd.read_csv('data/dataset.csv')
print(f"Rows: {len(df)}")
print(f"Target values: {df['target'].value_counts()}")
print(f"Missing values: {df.isnull().sum()}")

# Remove rows with missing target values
df = df.dropna(subset=['target'])
df.to_csv('data/cleaned_dataset.csv', index=False)

SHAP computation errors¶

Error:

Exception: TreeExplainer only supports the following model types: xgboost.Booster

Solution:

# Use compatible explainer
explainers:
  priority:
    - kernelshap # Model-agnostic alternative
  config:
    kernelshap:
      n_samples: 100

PDF generation issues¶

Error:

OSError: cannot load library 'pango-1.0-0'

Solution (Linux):

# Install required system libraries
sudo apt install libpango1.0-dev libcairo2-dev libgtk-3-dev

Solution (macOS):

# Install with Homebrew
brew install pango cairo gtk+3

Solution (Alternative):

# Use HTML output instead
# Modify report configuration:

report:
  output_format: html # Generate HTML instead of PDF

Performance issues¶

Slow execution¶

Problem: Audit takes longer than expected

Solutions:

Reduce Sample Sizes:

explainers:
  config:
    treeshap:
      max_samples: 100 # Default: 1000
    kernelshap:
      n_samples: 50 # Default: 500
      background_size: 50 # Default: 100

Enable Parallel Processing:

performance:
  n_jobs: -1 # Use all CPU cores

Use Simpler Models:

model:
  params:
    n_estimators: 50 # Reduce from 100
    max_depth: 3 # Reduce from 6

Skip Expensive Operations:

# Use dry run for config validation
glassalpha audit --config config.yaml --output test.pdf --dry-run

# Test with smaller datasets first
head -n 500 large_dataset.csv > test_dataset.csv

High memory usage¶

Problem: System runs out of memory during audit

Solutions:

Enable Memory Optimization:

performance:
  low_memory_mode: true
  n_jobs: 1 # Reduce parallelism

Process Data in Chunks:

# Pre-process large datasets
import pandas as pd
df = pd.read_csv('large_dataset.csv')
sample = df.sample(n=1000, random_state=42)  # Use sample for testing
sample.to_csv('sample_dataset.csv', index=False)

Optimize Explainer Settings:

explainers:
  config:
    treeshap:
      max_samples: 50 # Very small for memory constraints
    kernelshap:
      n_samples: 20
      background_size: 20

Large PDF files¶

Problem: Generated PDFs are too large

Solutions:

Optimize Report Configuration:

report:
  styling:
    optimize_size: true
    compress_images: true

Reduce Plot Resolution:

explainers:
  config:
    treeshap:
      plot_dpi: 150 # Default: 300

Exclude Optional Sections:

report:
  include_sections:
    - executive_summary
    - model_performance
    # Remove: local_explanations, detailed_plots

Debugging and diagnostics¶

Enable verbose logging¶

# Enable detailed logging
glassalpha --verbose audit --config config.yaml --output audit.pdf

# Check specific component status
export GLASSALPHA_LOG_LEVEL=DEBUG
glassalpha audit --config config.yaml --output audit.pdf

Explainer/model compatibility debugging¶

# Debug explainer availability
python -c "
from glassalpha.explain import select_explainer

# Test each model type
for model_type in ['xgboost', 'lightgbm', 'sklearn']:
    try:
        explainer = select_explainer(model_type)
        print(f'{model_type}: {explainer.__class__.__name__}')
    except Exception as e:
        print(f'{model_type}: ERROR - {e}')
"

Configuration debugging¶

# Debug configuration loading
python -c "
from glassalpha.config import load_config_from_file
config = load_config_from_file('your_config.yaml')
print('Loaded config:', config.model_dump())
"

Data loading debugging¶

# Debug data issues
python -c "
from glassalpha.data import TabularDataLoader
loader = TabularDataLoader()
data = loader.load('data/dataset.csv')
print('Shape:', data.shape)
print('Columns:', data.columns.tolist())
print('Dtypes:', data.dtypes.to_dict())
print('Missing:', data.isnull().sum().to_dict())
"

Getting help¶

Before requesting support¶

Check this troubleshooting guide
Run diagnostic commands shown above
Try with the German Credit example (known working configuration)
Enable verbose logging for detailed error information

Information to include in support requests¶

# System information
glassalpha --version
python --version
pip list | grep -E "(glassalpha|xgboost|lightgbm|shap|pandas)"

# Configuration (remove sensitive data)
cat your_config.yaml

# Complete error message
glassalpha --verbose audit --config config.yaml --output audit.pdf 2>&1

Support channels¶

GitHub Issues: https://github.com/GlassAlpha/glassalpha/issues
Documentation: Complete guides
Community: GitHub Discussions

Prevention best practices¶

Environment setup¶

# Always use virtual environments
python -m venv glassalpha-env
source glassalpha-env/bin/activate

# Keep dependencies updated
pip install --upgrade pip
pip install --upgrade glassalpha

Configuration management¶

# Always specify explicit seeds for reproducibility
reproducibility:
  random_seed: 42
# Use version control for configurations
# Git commit: "Update audit configuration for production"

# Validate configurations before use

glassalpha validate --config production.yaml --strict

Testing strategy¶

# Test with small datasets first
glassalpha audit --config config.yaml --output test.pdf --dry-run

# Test new configurations incrementally
# Start with minimal config, add features gradually

Monitoring and logging¶

# Enable logging for production
export GLASSALPHA_LOG_LEVEL=INFO

# Log all audit runs
glassalpha audit --config config.yaml --output audit.pdf 2>&1 | tee audit.log

This troubleshooting guide covers the most common issues encountered with GlassAlpha. If you encounter an issue not covered here, please check the GitHub issues or contact support with the diagnostic information requested above.

Troubleshooting guide¶

Quick diagnostics¶

CLI command issues¶

glassalpha: command not found¶

Installation issues¶

Python version compatibility¶

XGBoost installation issues (macOS)¶

Dependency conflicts¶

Missing system dependencies¶

Configuration errors¶

Missing required fields¶

Invalid file paths¶

Model type not found¶

Schema validation errors¶

Model and dependency issues¶

Model not available errors¶

Model loading failures¶

Fallback model not suitable¶

Explainer errors (SHAP library required)¶

Data issues¶

Data format mismatch in reasons/recourse commands¶

Dataset not found¶

Column not found¶

Protected attributes not found¶

Data format issues¶

Data type issues¶

Missing values¶

Target column issues¶

Insufficient data¶

Class imbalance¶

Runtime errors¶

Memory issues¶

Model training failures¶

SHAP computation errors¶

PDF generation issues¶

Performance issues¶

Slow execution¶

High memory usage¶

Large PDF files¶

Debugging and diagnostics¶

Enable verbose logging¶

Explainer/model compatibility debugging¶

Configuration debugging¶

Data loading debugging¶

Getting help¶

Before requesting support¶

Information to include in support requests¶

Support channels¶

Prevention best practices¶

Environment setup¶

Configuration management¶

Testing strategy¶

Monitoring and logging¶

`glassalpha: command not found`¶