sdmf 0.1.4

SDMF - Standard Data Management Framework

Standard Data Management Framework (SDMF)

A modular, scalable, and Python-based Data Management Framework designed to standardize data ingestion, validation, transformation, metadata handling, and storage across enterprise workflows.

This framework eliminates repetitive boilerplate and provides a consistent structure for building reliable, maintainable data pipelines.

About

Created and maintained by Harsh Handoo, Data Engineer, SDMF is designed to standardize common data movement patterns and reduce boilerplate in real-world Spark workloads.

SDMF (Standard Data Management Framework) is an open-source Spark-based data engineering framework built for reliable, production-grade data pipelines. It focuses on schema enforcement, incremental processing, and SCD Type-2 handling using Delta Lake.

Features

• Modular Design – Plug-and-play components for ingestion, validation, transformation, and storage.
• Schema Alignment & Partitioning – Built-in support for CDC (Change Data Capture) and MERGE operations.
• Metadata Management – Centralized handling of feed specifications and lineage.
• Scalable – Works seamlessly with Spark, Delta Lake, and distributed environments like Databricks.
• Logging & Monitoring – Custom logging with retention and rotation policies.

Installation

pip install sdmf

Requirements

Cluster Resources (Typical)

Workload	Minimum	Recommended
Local development	4 vCPU, 8 GB RAM	8 vCPU, 16 GB RAM
Small datasets (<10M rows)	2 executors × 4 GB	4 executors × 8 GB
Medium datasets (10–100M rows)	4 executors × 8 GB	8 executors × 16 GB
Large datasets (>100M rows)	8+ executors × 16 GB	Cluster-specific tuning

Recommended Production Setup

• Linux-based Spark cluster
• Spark FAIR scheduler enabled
• Delta Lake tables stored on cloud object storage
• Versioned releases via PyPI + GitHub Releases

Storage

• Local filesystem (dev only)
• HDFS / ADLS / S3 / GCS (recommended)
• DBFS (Databricks)

Operating System

• Linux (recommended)
• macOS
• Windows (WSL recommended for local development)

⚠️ Production deployments are strongly recommended on Linux-based systems.

Note: This library is tested on databricks.

Usage

Prerequisites

• Dedicate a directory to SDMF. Example: /sdmf_dir/

• Setup config.ini file.

  [DEFAULT]
  outbound_directory_name=sdmf_outbound
  log_directory_name=sdmf_logs
  temp_log_location=/sdmf_dir/temp
  file_hunt_path=/sdmf_dir/
  log_retention_policy_in_days=7
  max_concurrent_batches=4
  
  [FILES]
  master_spec_name = master_specs.xlsx
  
  [LINEAGE_DIAGRAM]
  BOX_WIDTH=4.4
  BOX_HEIGHT=2.2
  X_GAP=2.0
  Y_GAP=2.5
  ROOT_GAP=2.0
  

• Setup Master Spec master_spec.xlsx (can be renamed in config) file.

  • feed_id
  • system_name
  • subsystem_name
  • category
  • sub_category
  • data_flow_direction
  • residing_layer
  • feed_name
  • feed_type
  • feed_specs
  • load_type
  • target_unity_catalog
  • target_schema_name
  • target_table_name
  • suggested_feed_name
  • parallelism_group_number
  • parent_feed_id
  • is_active

• Feed Spec JSON

  {
      "primary_key": "col1",
      "composite_key": [],
      "partition_keys": [],
      "vacuum_hours": 168,
      "source_table_name": "test.test",
      "selection_query":null,
      "selection_schema": {
          "type": "struct",
          "fields": [
              {
                  "name": "col1",
                  "type": "string",
                  "nullable": true,
                  "metadata": {
                      "comment": "test"
                  }
              },
              {
                  "name": "col2",
                  "type": "string",
                  "nullable": true,
                  "metadata": {
                      "comment": "test"
                  }
              },
              {
                  "name": "col3",
                  "type": "string",
                  "nullable": true,
                  "metadata": {
                      "comment": "test"
                  }
              },
              {
                  "name": "col4",
                  "type": "string",
                  "nullable": true,
                  "metadata": {
                      "comment": "test"
                  }
              }
          ]
      },
      "standard_checks": [
          {
              "check_sequence": [
                  "_check_primary_key"
              ],
              "column_name": "col1",
              "threshold": 0
          },
          {
              "check_sequence": [
                  "_check_nulls"
              ],
              "column_name": "col2",
              "threshold": 0
          }
      ],
      "comprehensive_checks": [
          {
              "check_name": "Some unique check name",
              "query": "Select 1;",
              "severity": "WARNING",
              "threshold": 0,
              "load_stage": "PRE_LOAD",
              "dependency_dataset": []
          },
          {
              "check_name": "Some unique check name 1",
              "query": "Select 1;",
              "severity": "WARNING",
              "threshold": 0,
              "load_stage": "PRE_LOAD",
              "dependency_dataset": []
          },
          {
              "check_name": "Some unique check name 2",
              "query": "Select 1;",
              "severity": "WARNING",
              "threshold": 0,
              "load_stage": "PRE_LOAD",
              "dependency_dataset": []
          },
          {
              "check_name": "Some unique check name 3",
              "query": "Select 1;",
              "severity": "WARNING",
              "threshold": 0,
              "load_stage": "POST_LOAD",
              "dependency_dataset": [
                  "demo.customers"
              ]
          }
      ]
  }
  

• Ensure Spark FAIR scheduler is enabled.

  #!/bin/bash
  
  echo "Configuring Spark FAIR scheduler..."
  
  cat <<EOF >> /databricks/spark/conf/spark-defaults.conf
  spark.scheduler.mode FAIR
  EOF
  
  echo "Spark FAIR scheduler enabled."
  

  from pyspark.sql import SparkSession
  
  spark = (
      SparkSession.builder
          .appName("SDMF")
          .config("spark.scheduler.mode", "FAIR")
          .getOrCreate()
  )
  

Execution

import configparser
from sdmf import Orchestrator

spark # available spark session

cfg = configparser.ConfigParser()
cfg.read("/sdmf_dir/config.ini")
myOrchestrator = Orchestrator(spark, config=cfg)
myOrchestrator.run()

Logging

• Logs are first written to specified log directory in config.ini.