FOUNDATIONS OF

BLOCKCHAIN TECHNOLOGY
BCSE324L

Dr. Malathi D.
 FOUNDATIONS OF BLOCKCHAIN TECHNOLOGY

Course Objectives

• To understand building blocks of Blockchain.

• To significance of Distributed Ledger Technology and Smart Contract.
• To exploit applications of Blockchain in real world scenarios and their impacts.

Expected Outcomes

• Understand Blockchain ecosystem and its services in real world sceneries

• Apply and Analyze the requirement of Distributed Ledger Technology and Smart Contract
• Design and Demonstrate end-to-end decentralized applications
• Acquaint the protocol and assess their computational requirements

16-04-2025 Dr. Malathi D., SCOPE–VIT 2

Module – 7

High Performance Computing

16-04-2025 Dr. Malathi D., SCOPE–VIT 3

 High Performance Computing
• Integrity of High Performance Systems
• Data Provenance
• Cluster Construction and Deployment
• Mock Workload
• Blockchain Software Evaluation
• Blockchain storage of Integrity Data

16-04-2025 Dr. Malathi D., SCOPE–VIT 4

 High Performance Computing
• High Performance Computing (HPC) refers to the use of supercomputers and parallel processing
techniques to solve complex computational problems that are beyond the capabilities of traditional
computers.
• HPC systems enable researchers, engineers, and analysts to perform simulations, analyze massive
datasets, and run sophisticated algorithms at high speed and scale.

Key Components of HPC Systems

• Compute Nodes
• The backbone of an HPC system, comprising multiple processors (CPUs) or GPUs for executing
tasks.
• Each node may have multiple cores for parallel computation.
• High-Speed Interconnect
• Connects compute nodes to enable efficient communication.
• Examples: InfiniBand, Ethernet, Omni-Path.
• Storage
• Fast and scalable storage systems for handling large datasets.
• Includes parallel file systems like Lustre, GPFS, or BeeGFS.
16-04-2025 Dr. Malathi D., SCOPE–VIT 5
 High Performance Computing
• Software and Middleware
• Operating systems, job schedulers, and libraries that facilitate task management and parallel
processing.
• Common job schedulers: **Slurm**, **PBS Pro**, **HTCondor**.
• Cooling Systems
• Advanced cooling mechanisms to prevent overheating of dense compute clusters.
• Network and Power
• Redundant and efficient power supplies with high-speed network capabilities for seamless
operations.
Key Characteristics of HPC
• Parallel Processing: Tasks are divided into smaller subtasks and executed simultaneously across multiple
cores or nodes.
• Scalability: Ability to add more resources (nodes, storage, etc.) to meet increasing computational demands.
• High-Speed Networking: Enables low-latency communication between nodes to support distributed
computing.
• Fault Tolerance: Systems are designed to handle hardware or software failures without disrupting
workloads.
16-04-2025 Dr. Malathi D., SCOPE–VIT 6
 High Performance Computing
Applications of HPC
• Scientific Research
• Simulations for weather prediction, climate modeling, and astrophysics.
• Computational chemistry, genomics, and drug discovery.
• Engineering and Manufacturing
• Computational fluid dynamics (CFD) for aircraft design.
• Structural analysis and materials science simulations.
• Data Analysis
• Big data analytics in fields like finance, healthcare, and business intelligence.
• Machine learning and AI model training.
• Energy
• Oil and gas exploration using seismic data analysis.
• Nuclear and renewable energy simulations.
• National Security
• Cryptography, surveillance, and defense simulations.

16-04-2025 Dr. Malathi D., SCOPE–VIT 7

 High Performance Computing
Technologies Enabling HPC
• Processors
• Multi-core CPUs and specialized accelerators like GPUs (NVIDIA, AMD) and TPUs.
• ARM-based processors (e.g., AWS Graviton).
• Memory and Storage
• High-bandwidth memory (HBM) for fast data access.
• Non-volatile memory express (NVMe) storage for high-speed I/O operations.
• Parallel Programming Models
• MPI (Message Passing Interface) for distributed computing.
• OpenMP for shared-memory programming.
• CUDA for GPU programming.
• Cloud HPC
• Cloud providers offering scalable HPC resources, such as:
• AWS HPC
• Azure HPC
• Google Cloud HPC
• Quantum Computing
• Emerging as a future extension of HPC for solving quantum-based problems.
16-04-2025 Dr. Malathi D., SCOPE–VIT 8
 High Performance Computing
Challenges in HPC
• Energy Consumption: HPC systems consume significant power, requiring efficient energy management.
• Scalability: Ensuring linear scalability as more resources are added.
• Software Optimization: Writing and optimizing software to leverage parallelism effectively.
• Cost: High initial setup costs for hardware, software, and infrastructure.
• Data Management: Efficiently storing, processing, and moving massive datasets.

HPC Ecosystem and Tools

• HPC Frameworks
• Apache Spark: Scalable data analytics for large-scale workloads.
• Hadoop: Distributed processing for big data.
• Cluster Management
• Slurm: Open-source job scheduling and resource management.
• Kubernetes (with MPI): For containerized HPC applications.
• Development Tools
• Intel Parallel Studio: For optimizing parallel applications.
• NVIDIA CUDA Toolkit: For GPU acceleration.
• PGI Compilers: Optimized compilers for HPC.
16-04-2025 Dr. Malathi D., SCOPE–VIT 9
 High Performance Computing
• Monitoring Tools
• Ganglia: Monitoring and metrics collection for clusters.
• Prometheus and Grafana: Metrics collection and visualization.

Future Trends in HPC

• Exascale Computing
• Systems capable of performing 1 exaFLOP (10¹⁸ floating-point operations per second).
• Examples: Frontier (US), Aurora, El Capitan.
• AI and HPC Convergence
• HPC systems tailored for AI and ML workloads.
• Optimized AI frameworks for HPC environments.
• Edge HPC
• Distributed HPC resources closer to data sources for real-time analytics.
• Green Computing
• Emphasis on energy efficiency with sustainable and carbon-neutral designs.
• Quantum-HPC Hybrid Systems
• Integration of classical HPC with quantum computing for specific tasks.
16-04-2025 Dr. Malathi D., SCOPE–VIT 10
 Integrity of High Performance Systems
• Ensuring the integrity of High-Performance Computing (HPC) systems involves maintaining their
operational reliability, security, and accuracy while safeguarding data, software, and hardware from
corruption, tampering, or unauthorized alterations.
• Integrity is essential in HPC environments where errors or breaches can compromise critical
computations or sensitive data.

Key Aspects of HPC Integrity

• Data Integrity: Ensuring data stored, processed, or transmitted remains accurate, complete, and
unaltered. Includes mechanisms for error detection and correction.
• System Integrity: Protecting the HPC infrastructure, including compute nodes, storage, and network,
from unauthorized changes or failures.
• Software Integrity: Guaranteeing that applications and operating systems are not tampered with or
corrupted during execution.
• Process Integrity: Ensuring tasks running on HPC systems execute as intended without interference or
errors.

16-04-2025 Dr. Malathi D., SCOPE–VIT 11

 Integrity of High Performance Systems
Threats to HPC Integrity
• Hardware Failures: Bit flips caused by cosmic rays or hardware malfunctions can lead to data corruption.
• Software Bugs: Errors in application code or system software can introduce inaccuracies.
• Malware or Cyberattacks: Unauthorized access, injection attacks, or malicious code can compromise HPC
systems.
• Human Error: Misconfigurations or accidental overwrites can affect data and processes.
• Network Issues: Packet loss or tampering during data transfer can compromise results.

Maintaining the Integrity of HPC Systems

Data Integrity Measures
1. Checksums and Hashing
• Use checksums or cryptographic hashes (e.g., SHA-256) to verify data integrity during transfers or at
rest.
• Examples: Verifying file integrity before job execution.
2. Error-Correcting Codes (ECC)
• Use ECC memory to detect and correct single-bit errors in memory.
• Reduces risks of data corruption from hardware issues.
16-04-2025 Dr. Malathi D., SCOPE–VIT 12
 Integrity of High Performance Systems
3. Data Backup and Redundancy
• Maintain regular, secure backups.
• Use RAID configurations or distributed file systems (e.g., Lustre, GPFS) to ensure redundancy.

System Integrity Measures

1. Secure Boot
• Ensure HPC nodes boot only trusted firmware and operating systems.
• Use technologies like Trusted Platform Modules (TPMs).
2. Access Control
• Enforce role-based access control (RBAC) and the principle of least privilege (PoLP).
• Regularly review and update user permissions.
3. Patch Management
• Regularly update and patch system software to address vulnerabilities.

Software Integrity Measures

1. Code Signing
• Use digital signatures to verify the authenticity of software and scripts running on HPC systems.
2. Dependency Management
• Monitor libraries and dependencies for vulnerabilities using tools like Snyk or Dependabot.
16-04-2025 Dr. Malathi D., SCOPE–VIT 13
 Integrity of High Performance Systems
3. Testing and Validation
• Use automated and manual testing to validate the correctness of application code.
Process Integrity Measures
1. Job Scheduling and Monitoring
• Use robust job schedulers (e.g., Slurm, PBS) to manage and monitor tasks.
• Detect and address anomalies during execution.
2. Audit Trails
• Maintain logs of system access, job executions, and data modifications to trace unauthorized changes.
3. Isolation
• Use containerization (e.g., Docker, Singularity) or virtual machines to isolate jobs and prevent
interference.
Network Integrity Measures
1. Encryption
• Use TLS/SSL to secure data in transit across HPC interconnects.
2. Intrusion Detection
• Deploy network intrusion detection systems (NIDS) to monitor and flag suspicious activity.
3. High-Speed Interconnect Monitoring
• Monitor HPC interconnects like InfiniBand for packet integrity and latency issues.
16-04-2025 Dr. Malathi D., SCOPE–VIT 14
 Integrity of High Performance Systems
Tools for Ensuring HPC Integrity

16-04-2025 Dr. Malathi D., SCOPE–VIT 15

 Integrity of High Performance Systems
Future Trends in HPC Integrity
• Quantum-Safe Integrity: As quantum computing emerges, ensuring cryptographic techniques are
resistant to quantum attacks.
• AI-Driven Monitoring: Use AI to predict and detect anomalies in HPC operations for proactive
integrity management.
• Blockchain: Applying blockchain for tamper-proof logging of computations and results.
• Self-Healing Systems: HPC systems that can automatically detect and recover from integrity breaches.

• Integrity in HPC systems is vital to ensuring that computations, data, and results are reliable and
trustworthy.
• By implementing robust mechanisms such as encryption, checksums, secure boot, and regular
monitoring, organizations can safeguard their HPC environments from threats and errors.

16-04-2025 Dr. Malathi D., SCOPE–VIT 16

 Data Provenance
• Understanding where your data comes from, how it’s processed, and where it goes.
• It’s more than just tracking data — it’s about ensuring the integrity, security, and transparency of your
data throughout its life cycle.

What is Data Provenance?

• Data provenance tracks the history of data: where it started, how it has been processed, and how it
moves across systems.
• Essentially, it answers two key questions: "Where did this data come from?" and "How has it changed
over time?“
• For organizations, tracking data provenance ensures that their data remains reliable and trustworthy,
which is especially crucial in sectors like healthcare, finance, and research, where data accuracy is vital.

16-04-2025 Dr. Malathi D., SCOPE–VIT 17

 Data Provenance
Types of Data Provenance
• Data provenance is usually categorized into two types:
• Backward Provenance (Retrospective): This tracks a dataset’s past, including its origin, changes,
and movements. It answers, "How did this data get here?“
• Forward Provenance (Prospective) : This forecasts how data will be handled in the future, often
included in workflow systems to predict future data movements and transformations.

Key Components of Data Provenance

• Data provenance can be broken down into four main components:
1. Data Lineage: Shows the data’s journey, detailing where it started and where it ended up. This is
especially important for regulatory compliance and transparency.
2. Data Source: Identifies the origin of the data—whether it’s a sensor, database, or external provider.
This adds credibility and traceability.
3. Data Transformation: Tracks changes made to data, such as cleaning, aggregating, or modifying it.
This ensures the data is consistent, and any errors can be traced back to a specific change.
4. Data Destination: Shows where the data is stored or used after processing, confirming that it reaches
the correct stakeholders or systems.
16-04-2025 Dr. Malathi D., SCOPE–VIT 18
 Data Provenance
Data Provenance vs. Data Lineage
• While both data provenance and data lineage focus on the history of data, they serve different purposes.
• Data lineage maps out the path data takes from its source to its destination, essentially showing a
roadmap of its journey.
• Data provenance goes further, also tracking any transformations the data undergoes and adding
context to help understand the full lifecycle of the data.

Data Provenance vs. Metadata

• While closely related, data provenance and metadata have distinct roles.
• Metadata provides basic information about the data itself, such as file size, creation date, or author.
• Data provenance, details how the data has been used, changed, and shared over time.

Standards and Protocols for Data Provenance

• Various standards ensure consistency in tracking data provenance:
• W3C PROV Standard: A standard for documenting data provenance on the web.
• ISO Standards: Widely used in scientific and research settings for managing data provenance.
• Adhering to these standards ensures interoperability between systems and promotes best practices.
16-04-2025 Dr. Malathi D., SCOPE–VIT 19
 Data Provenance
Techniques for Tracking Data Provenance
• There are various methods to track data provenance, from simple manual logs to sophisticated
automated systems:
1. Manual Documentation: Manually recording data movements and changes. While useful for small
datasets, this is error-prone and doesn’t scale well.
2. Automated Provenance Collection: Automated systems use tools and software to collect and record
data provenance information in real time, ensuring accuracy and efficiency. These systems can capture
detailed logs of data transformations, sources, and destinations with minimal human intervention.
3. Blockchain: Blockchain’s decentralized and immutable structure makes it useful for tracking data
provenance. By storing provenance information on a blockchain, organizations can ensure data
integrity and prevent tampering.
4. Metadata Management Systems: Systems that record metadata (descriptive details about data),
allowing organizations to track the data’s complete lifecycle.
5. Provenance-Aware Storage Systems: Storage systems that have built-in capabilities for recording
data provenance, ensuring all changes are documented. These systems are designed to support the
recording and tracking of data provenance as part of their core functionality.
16-04-2025 Dr. Malathi D., SCOPE–VIT 20
 Data Provenance
Benefits of Data Provenance
• Data Integrity: Provenance ensures that data has not been tampered with, helping to maintain trust in its
accuracy and quality.
• Regulatory Compliance: Many industries are subject to stringent regulations requiring transparency in
data handling. Provenance tracking helps organizations meet these requirements.
• Error Detection: With complete data lineage, it’s easier to identify where errors or inconsistencies
originated, leading to faster resolution of data quality issues.
• Enhanced Security: Identify potential security breaches by highlighting any unauthorized changes to
data.

Challenges of Data Provenance

• Scalability: As data grows in volume, tracking its provenance can become resource-intensive, Managing
and storing detailed provenance information for large datasets requires advanced tools and infrastructure.
• Privacy Concerns: In certain cases, provenance data may reveal sensitive information about the data’s
source or handling, raising privacy issues.
• Complexity: Implementing a comprehensive provenance system can be complex, especially when dealing
with multiple systems and data sources that don’t natively support provenance tracking.
16-04-2025 Dr. Malathi D., SCOPE–VIT 21
 Data Provenance
Real-World Applications of Data Provenance
Healthcare
• Imagine a health care provider conducting a clinical trial for a new medication. They gather data from
patients — such as blood pressure, temperature, and other vital signs — from multiple sources (e.g.,
monitoring devices, lab results, patient surveys).
• Provenance tracking ensures accurate recording of this data, from the initial collection to the various
transformations it undergoes. If researchers find an anomaly in the results, the provenance system
helps them trace it back to a specific data source or transformation stage, allowing them to quickly
identify and rectify errors.
Finance
• In the finance industry, banks often rely on complex data systems to track customer transactions,
manage portfolios, and generate reports for regulatory compliance.
• Example: when auditing a financial transactions, an auditor needs to trace each transaction back to its
origin, ensuring that all data — from the initial entry to the final report — is accurate and unaltered.
• With a data provenance system in place, every transformation (e.g., currency conversion, interest
calculation) and movement of data is tracked.
16-04-2025 Dr. Malathi D., SCOPE–VIT 22
 Data Provenance
• If discrepancies arise in a report, the system identifies where incorrect data may have been introduced,
such as during a specific stage of data aggregation.
• This process helps ensure that the financial institution complies with regulations like the Sarbanes-
Oxley Act, which requires thorough documentation of financial data.
Scientific Research
• In scientific research, data provenance is key to ensuring the validity and reproducibility of
experiments.
• Consider a research study on climate change that aggregates data from various sources, such as satellite
imagery, weather station measurements, and historical climate records.
• Provenance tracking would document how raw data from these sources is collected, processed (e.g.,
normalizing temperature readings from different formats), and analyzed.
• If the final research report identifies a trend that conflicts with previous studies, provenance can help
verify that all data used was correct and processed properly.
• Additionally, when sharing data with other researchers, a detailed provenance record ensures that they
can replicate the study with full confidence in the data’s integrity, thus supporting reproducibility , one
of the cornerstones of scientific research.
16-04-2025 Dr. Malathi D., SCOPE–VIT 23
 Cluster Construction and Deployment
• Building and deploying a cluster for HPC involves careful planning, selecting the right hardware and
software components, and setting up infrastructure to support the desired computational tasks.
• Clusters are typically a collection of interconnected computers (nodes) that work together as a single
system to perform large-scale computations.

Phases of Cluster Construction and Deployment

• Planning and Design
• Hardware Assembly
• Software Configuration
• Network Setup
• Testing and Optimization
• Deployment and Maintenance

16-04-2025 Dr. Malathi D., SCOPE–VIT 24

 Cluster Construction and Deployment
Planning and Design
• Determine Requirements
• Use Case: Identify the type of workloads (e.g., simulations, data analytics, machine learning).
• Performance Needs: CPU/GPU requirements, memory size, storage capacity, and network speed.
• Scalability: Plan for future expansions of the cluster.
• Choose the Cluster Architecture
• Single Head Node: Manages job scheduling and resource allocation.
• Compute Nodes: Perform the computations.
• Storage Nodes: Centralized or distributed storage for datasets.
• Interconnect: High-speed networking technology (e.g., InfiniBand, Ethernet).
• Select the Operating System
• Popular choices include Linux distributions (CentOS, Rocky Linux, Ubuntu).

Hardware Assembly
• Core Hardware Components
1. Compute Nodes:
• Multi-core CPUs (e.g., AMD EPYC, Intel Xeon) or GPUs (e.g., NVIDIA A100, AMD Instinct).
16-04-2025 Dr. Malathi D., SCOPE–VIT 25
 Cluster Construction and Deployment
• High-bandwidth memory (DDR5, HBM).
2. Storage:
• Fast storage (SSD, NVMe) for active computations.
• High-capacity HDDs for archival purposes.
• Parallel file systems like Lustre, BeeGFS, or GPFS.
3. Networking:
• High-speed interconnects (e.g., 100Gbps InfiniBand).
• Network switches with sufficient ports.
4. Cooling:
• Rack-based liquid cooling or airflow cooling to handle thermal loads.
5. Power Supply:
• Uninterruptible power supply (UPS) to ensure uptime.
• Infrastructure
• Rack Mounting: Install nodes in racks for space-efficient management.
• Cabling: Connect power and network cables systematically.

16-04-2025 Dr. Malathi D., SCOPE–VIT 26

 Cluster Construction and Deployment
Software Configuration
• Operating System Installation
• Install the OS on all nodes, typically a lightweight and performance-optimized Linux distribution.
• Use tools like PXE Boot or disk cloning to automate installations.
• Cluster Management Software
• Head Node Configuration:
• Acts as the control center for job scheduling, monitoring, and resource management.
• Popular tools for cluster management:
• Slurm: Open-source workload manager.
• PBS Pro: Job scheduling and management.
• HTCondor: High-throughput job scheduling.
• Middleware
• MPI (Message Passing Interface):
• Libraries like OpenMPI or MPICH for distributed parallel computing.
• Cluster Libraries:
• Install optimized mathematical libraries (e.g., BLAS, LAPACK, FFTW).
• Containerization (Optional)
• Tools like Singularity or Docker to run applications in isolated environments.
16-04-2025 Dr. Malathi D., SCOPE–VIT 27
 Cluster Construction and Deployment
Network Setup
• Node Communication
• Configure the network to ensure low-latency communication between nodes.
• Set up private IP ranges for internal node communication.
• Implement DNS or hostname resolution for easier management.
• File Sharing
• Use distributed file systems (e.g., NFS, Lustre) for shared access to data.
• Set up secure data transfer protocols (e.g., SCP, Rsync).
• Security
• Configure firewalls to restrict access to the cluster.
• Use SSH keys for secure and automated access to nodes.

Testing and Optimization

• Initial Tests
• Verify hardware health using diagnostics tools (e.g., memtest86 for memory, iperf for network
speed).
• Check connectivity between nodes with tools like ping or traceroute.
16-04-2025 Dr. Malathi D., SCOPE–VIT 28
 Cluster Construction and Deployment
• Benchmarking
• Test system performance with benchmarks specific to your workload:
• LINPACK: Measures floating-point computing power.
• Iozone: Tests file system performance.
• HPCG: Tests HPC-specific performance.
• Optimization
• Fine-tune the job scheduler configurations for workload balance.
• Optimize network settings for reduced latency (e.g., enabling jumbo frames for Ethernet).
• Apply compiler optimizations for software performance (e.g., GCC, Intel, or LLVM compilers).

Deployment and Maintenance

• Production Deployment
• Deploy the cluster into its production environment.
• Set up job queues, resource allocation policies, and access controls.
• Monitoring and Logging
• Tools like Ganglia, Nagios, or Prometheus for monitoring system health and performance.
• Centralized logging for diagnostics and audit trails.
16-04-2025 Dr. Malathi D., SCOPE–VIT 29
 Cluster Construction and Deployment
• Regular Maintenance
• Schedule regular updates for software, firmware, and OS.
• Clean hardware and inspect cabling periodically.
• Conduct periodic security audits.
• User Support
• Provide training or documentation for researchers and engineers using the cluster.

Cluster Deployment Workflow Summary

16-04-2025 Dr. Malathi D., SCOPE–VIT 30

 Cluster Construction and Deployment
Tools for Cluster Management

• By following this structured approach, you can construct and deploy a scalable and efficient HPC
cluster tailored to your computational needs.

16-04-2025 Dr. Malathi D., SCOPE–VIT 31

 Mock Workload
• A mock workload simulates real-world computational tasks on a High-Performance Computing (HPC)
cluster to test and evaluate system performance, stability, and scalability.
• Mock workloads mimic typical user tasks, allowing administrators to validate cluster configurations,
optimize performance, and identify bottlenecks without requiring actual user data or applications.

When to Use Mock Workloads

• During Cluster Testing: Before deploying an HPC cluster, mock workloads validate its readiness to
handle real tasks.
• Performance Benchmarking: To assess compute, memory, network, and storage capabilities.
• Optimization: To tune hardware or software settings for better performance.
• Fault Testing: To identify system behavior under load or stress scenarios.
• Training and Demonstration: For user training or showcasing system capabilities without involving
real projects.

16-04-2025 Dr. Malathi D., SCOPE–VIT 32

 Mock Workload
Types of Mock Workloads
• CPU-Intensive
• Tasks that stress processor cores, e.g., numerical computations or matrix multiplications.
• Example tools: LINPACK, HPCC.
• Memory-Intensive
• Simulations or applications requiring large memory bandwidth or capacity.
• Example: Finite Element Analysis (FEA) models.
• I/O-Intensive
• Simulates high read/write operations to test storage and file systems.
• Example tools: Iozone, FIO.
• Network-Intensive
• Tests interconnect speeds and latency between nodes.
• Example tools: iperf, OSU Micro-Benchmarks.
• Mixed Workloads
• Combine different types of computational, memory, & I/O loads to simulate real-world scenarios.
• Example: A job involving molecular dynamics with frequent I/O and inter-node communication.
16-04-2025 Dr. Malathi D., SCOPE–VIT 33
 Mock Workload
Setting Up a Mock Workload
• Define the Test Objectives
• Determine what aspect of the cluster to evaluate (e.g., CPU, memory, I/O, or network).
• Set performance targets and thresholds.
• Select or Create Workload Simulators
• Use workload simulation tools (e.g., LINPACK, SLURM test jobs).
• Write custom scripts to mimic application behavior.
• Allocate Resources
• Specify the number of nodes, cores, memory, and storage for the mock task.
• Schedule jobs via the job scheduler (e.g., Slurm, PBS).
• Monitor During Execution
• Use monitoring tools (e.g., Prometheus, Ganglia) to track resource usage, temperatures, and
performance metrics.
• Analyze Results
• Compare observed performance with expectations.
• Identify bottlenecks or inefficiencies.
16-04-2025 Dr. Malathi D., SCOPE–VIT 34
 Mock Workload
Example Mock Workload
• Scenario: Testing an HPC cluster for scientific simulations.
• Objective: Evaluate the cluster’s ability to handle CPU-intensive tasks with significant inter-node
communication.
• Steps:
• Select a Workload
• Use High-Performance Linpack (HPL) to solve large systems of linear equations.
• Install Required Tools
• Install MPI (e.g., OpenMPI) and HPL binaries on the cluster.
• Prepare the Job Script
• Create a Slurm batch script to run the HPL benchmark:

16-04-2025 Dr. Malathi D., SCOPE–VIT 35

 Mock Workload
• Submit the Job
• Use `sbatch` to submit the job to the scheduler:
• Monitor
• Check resource usage during the run using `squeue` or monitoring dashboards.

• Tools for Mock Workloads

16-04-2025 Dr. Malathi D., SCOPE–VIT 36

 Mock Workload
Interpreting Results
• Benchmark Scores: Compare results against standard benchmarks for similar systems.
• Resource Utilization: Identify resource underutilization or bottlenecks (e.g., memory usage, CPU idle
times).
• Error Detection: Look for job failures, hardware errors, or configuration issues.

Best Practices
• Realistic Workload Representation: Ensure the mock workload closely resembles actual user tasks.
• Isolated Environment: Run tests in isolation to avoid interference from other processes.
• Iterative Testing: Conduct multiple tests with varying configurations to cover edge cases.
• Comprehensive Monitoring: Log and analyze all aspects, including CPU/GPU utilization, I/O rates,
and network latency.

• By running mock workloads, you can validate the operational efficiency and reliability of an HPC
cluster before it is put into production, ensuring it meets user requirements and performs optimally.

16-04-2025 Dr. Malathi D., SCOPE–VIT 37

 Blockchain Software Evaluation
• Blockchain technology has gained significant attention across various industries for its ability to offer
secure, decentralized, and transparent systems.
• Evaluating blockchain software effectively is critical for selecting the right platform based on specific
use cases, performance requirements, and the system's overall suitability.
• This evaluation process includes several factors, such as performance, security, scalability, and ease of
integration. Below is a structured approach to evaluating blockchain software:
Key Factors to Evaluate Blockchain Software
• Consensus Mechanism
• Security Features
• Scalability
• Performance and Speed
• Interoperability and Integration
• Governance Model
• Cost and Licensing
• Developer Community and Ecosystem
• Privacy Features
• Compliance and Regulatory Support
16-04-2025 Dr. Malathi D., SCOPE–VIT 38
 Blockchain Software Evaluation
1. Consensus Mechanism
• The consensus mechanism is the algorithm used to validate transactions and add them to the
blockchain.
• Common mechanisms include:
• Proof of Work (PoW): Used by Bitcoin; requires computational power to validate transactions.
• Proof of Stake (PoS): Used by Ethereum 2.0; relies on validators who lock up coins to validate
blocks.
• Delegated Proof of Stake (DPoS): Delegates voting power to a few trusted nodes.
• Practical Byzantine Fault Tolerance (PBFT): Used for permissioned blockchains, ensures
resilience even when some nodes are faulty.
• Evaluation Criteria:
• Energy efficiency (PoW vs. PoS)
• Transaction speed and finality
• Level of decentralization
• Fault tolerance and security

16-04-2025 Dr. Malathi D., SCOPE–VIT 39

 Blockchain Software Evaluation
2. Security Features
• Blockchain security is crucial to protect the integrity of data and prevent unauthorized access. Key
security features include:
• Cryptographic Algorithms: Ensure data encryption and secure communication.
• Node Validation and Authentication: Mechanisms to prevent unauthorized nodes from joining
the network.
• Immutability: Once recorded, transactions should be irreversible and tamper-proof.
• Smart Contract Auditing: Tools and techniques for verifying smart contract code to avoid
vulnerabilities (e.g., reentrancy attacks, overflow errors).
• Evaluation Criteria
• Level of encryption used (e.g., RSA, ECC)
• Protection against Sybil attacks, double-spending, and 51% attacks
• Support for multi-signature wallets and hardware wallets
• Auditability of transactions

16-04-2025 Dr. Malathi D., SCOPE–VIT 40

 Blockchain Software Evaluation
3. Scalability
• Scalability is a critical factor in determining whether the blockchain platform can handle increasing transaction
loads and maintain its performance.
• On-Chain Scaling: Increasing block size or transaction speed.
• Off-Chain Scaling: Layer 2 solutions like the Lightning Network or state channels to improve
throughput without congesting the main chain.
• Sharding: Splitting the blockchain into smaller parts (shards) that operate concurrently to increase
capacity.
• Evaluation Criteria
• Transaction throughput (transactions per second)
• Block size and time to confirm transactions
• Latency and how the system performs with increased demand
• Solutions for addressing blockchain bloat (e.g., pruning)

4. Performance and Speed

• Performance refers to the transaction throughput, latency, and overall speed at which the blockchain can
process transactions and execute smart contracts.

16-04-2025 Dr. Malathi D., SCOPE–VIT 41

 Blockchain Software Evaluation
• Transaction Throughput (TPS): The number of transactions that can be processed per second.
• Block Confirmation Time: How quickly a transaction is confirmed once added to the blockchain.
• Smart Contract Execution Time: Speed at which decentralized applications (dApps) are executed.
• Evaluation Criteria
• Transaction speed and block finality
• Latency under different loads
• Load testing and stress testing results
• Smart contract execution time

5. Interoperability and Integration

• For a blockchain to be useful in a multi-chain world, interoperability is essential. Blockchain platforms
should facilitate easy communication with other blockchains or legacy systems.
• Cross-Chain Communication: Protocols like Polkadot or Cosmos allow multiple blockchains to
interact.
• API Support: Provides standard protocols for integrating the blockchain with external systems,
applications, and databases.
• Oracles: Data feeds to integrate real-world data into smart contracts (e.g., Chainlink).
16-04-2025 Dr. Malathi D., SCOPE–VIT 42
 Blockchain Software Evaluation
• Evaluation Criteria
• Availability of cross-chain bridges or protocols
• APIs for external system integration
• Compatibility with existing legacy systems
• Use of oracles for external data feeds
6. Governance Model
• The governance model determines how decisions about updates, protocol changes, and network rules
are made. Blockchain networks can be permissioned (centralized) or permissionless (decentralized).
• On-Chain Governance: Involves all participants in the decision-making process, e.g., Ethereum’s
Improvement Proposals (EIPs).
• Off-Chain Governance: Governance happens off-chain, often by developers or a small group of
entities.
• Evaluation Criteria
• Decentralization of decision-making
• Transparency of the governance process
• Mechanisms for proposing and voting on changes
• Community involvement
16-04-2025 Dr. Malathi D., SCOPE–VIT 43
 Blockchain Software Evaluation
7. Cost and Licensing
• The cost of using a blockchain platform can significantly affect its viability for businesses. This includes setup
costs, transaction fees, and licensing terms.
• Transaction Fees: Fees paid to miners or validators for processing transactions.
• Licensing Model: Whether the blockchain is open-source or proprietary and the associated costs.
• Operational Costs: Energy consumption, server costs, etc.
• Evaluation Criteria
• Transaction fees and scalability of the fee model
• Cost of running nodes or validators
• Licensing model (open-source vs. commercial)
• Infrastructure and energy costs
8. Developer Community and Ecosystem
• A strong and active developer community helps ensure continuous innovation, bug fixes, and adoption. A
large ecosystem of tools, libraries, and frameworks supports easy development and integration.
• Evaluation Criteria
• Size and activity of the developer community
• Availability of documentation, tutorials, and resources
• Number of dApps or services built on the platform
• Ecosystem maturity and available tools
16-04-2025 Dr. Malathi D., SCOPE–VIT 44
 Blockchain Software Evaluation
9. Privacy Features
• Blockchain networks can be public or private. Privacy features such as encryption and anonymity are
critical for use cases involving sensitive data.
• Private Blockchains: Permissioned blockchains with controlled access.
• Zero-Knowledge Proofs (ZKPs): Cryptographic methods that ensure data is valid without
revealing the actual data.
• Ring Signatures: Used for anonymous transactions in networks like Monero.
• Evaluation Criteria
• Privacy guarantees (e.g., confidential transactions)
• Level of anonymity and data protection
• Use of ZKPs or other privacy-enhancing technologies
• Compliance with privacy regulations (e.g., GDPR)

10. Compliance and Regulatory Support

• Given the emerging regulatory environment surrounding blockchain technology, it is essential to ensure
the platform can comply with legal requirements such as anti-money laundering (AML) and know-
your-customer (KYC) regulations.
16-04-2025 Dr. Malathi D., SCOPE–VIT 45
 Blockchain Software Evaluation
• Evaluation Criteria
• Built-in tools for KYC/AML compliance
• Integration with regulatory reporting systems
• Legal standing of the blockchain (e.g., support for data sovereignty)

• Evaluating blockchain software requires a detailed analysis of various factors such as consensus
mechanisms, security, scalability, and developer support.
• By aligning the platform’s strengths with the specific needs of your application—whether that be speed,
privacy, cost, or integration—organizations can select the most appropriate blockchain software for
their use case.

16-04-2025 Dr. Malathi D., SCOPE–VIT 46

 Blockchain Software Evaluation
Evaluating Blockchain Platforms: Comparison Table

16-04-2025 Dr. Malathi D., SCOPE–VIT 47

 Blockchain storage of Integrity Data

16-04-2025 Dr. Malathi D., SCOPE–VIT 48

 Blockchain storage of Integrity Data
• Storing integrity data on a blockchain is a robust approach to ensure the immutability, transparency,
and verifiability of critical information.

What is Integrity Data?

• Integrity data refers to information used to validate the authenticity and integrity of a system, record,
or process.
• This can include:-
• Hashes of files to verify that data hasn’t been tampered with.
• Audit trails for compliance and regulatory purposes.
• Digital signatures to authenticate users or transactions.
• Event logs to track changes in a system.

Blockchain as a Storage Medium for Integrity Data

• Blockchain is well-suited for storing integrity data due to its decentralized and immutable nature.

16-04-2025 Dr. Malathi D., SCOPE–VIT 49

 Blockchain storage of Integrity Data
How It Works
• Hashing Data
• Instead of storing raw data, a cryptographic hash (e.g., SHA-256) is generated for the data.
• This reduces storage requirements and ensures privacy.
• The hash represents a unique fingerprint of the original data.
• Storing Hashes on Blockchain
• The hash is written into a blockchain transaction.
• This action timestamps the hash and ensures its immutability.
• Data Verification
• To verify the integrity later, the original data is rehashed, and the result is compared with the stored hash
on the blockchain. If the hashes match, the data is intact.

Benefits of Using Blockchain for Integrity Data

• Immutability: Data recorded on a blockchain cannot be altered retroactively.
• Transparency: All participants in the blockchain network can view the recorded hashes.
• Decentralization: Eliminates single points of failure, enhancing security and trust.
• Traceability: Provides an auditable trail of changes, crucial for regulatory compliance.
• Tamper Resistance: Data is cryptographically secured, preventing unauthorized modifications.
16-04-2025 Dr. Malathi D., SCOPE–VIT 50
 Blockchain storage of Integrity Data
Use Cases
• Supply Chain Management: Ensuring the authenticity of products through immutable records of origin
and handling.
• Digital Rights Management: Verifying ownership and preventing piracy.
• Healthcare: Storing patient consent forms or medical records' hashes to ensure data integrity.
• Legal and Compliance: Maintaining secure and tamper-proof audit logs for regulatory purposes.
• Cybersecurity: Storing integrity checksums for software updates to prevent malware injection.

Storage Considerations
• While blockchain provides integrity, it's not ideal for storing large volumes of data.
• Instead:
• On-Chain Storage: Store only critical integrity data, such as hashes or metadata.
• Off-Chain Storage: Use traditional databases or decentralized file systems (e.g., IPFS, Filecoin) for
raw data, while the blockchain holds the hash.

16-04-2025 Dr. Malathi D., SCOPE–VIT 51

 Blockchain storage of Integrity Data
Example Workflow
• A company generates a hash of their financial report.
• The hash is stored on a public blockchain with a timestamp.
• Years later, an auditor retrieves the financial report, rehashes it, and compares it to the blockchain-
stored hash.
• A match confirms that the report hasn’t been altered.
Challenges and Considerations
• Scalability: High transaction volumes can strain blockchain networks.
• Cost: Blockchain transactions often come with fees (e.g., gas fees on Ethereum).
• Privacy Concerns: While hashes are secure, public blockchains may expose metadata.
Emerging Technologies for Optimization
• Layer 2 Solutions: Reduce costs and improve transaction speeds (e.g., Optimistic Rollups, zk-Rollups).
• Permissioned Blockchains: Limit access to authorized participants for privacy-sensitive applications.
• Interoperability Protocols: Enable smooth integration with off-chain systems and other blockchains.

• By leveraging blockchain technology for integrity data storage, organizations can ensure data
remains secure, verifiable, and resistant to tampering, which is vital in many modern applications.
16-04-2025 Dr. Malathi D., SCOPE–VIT 52
 Blockchain storage of Integrity Data
How Blockchain is monitoring and keeping our data secure?
1) Immutable Data Storage
• Data, once recorded, cannot be altered or deleted without the approval of network participants.
• Achieved by linking each block (or record) in the blockchain to the previous one through
cryptographic hashes.
• When data is added to a blockchain, it creates a permanent, tamper-proof record.
• Provides a secure way to store sensitive information—such as transaction histories or customer data—
ensuring that records remain accurate and cannot be illegally altered.
• Prevent data tampering and ensures data integrity, as any attempt to change a past record would
require changing every subsequent block, which would require immense computational power and
consensus from network participants.

2) Decentralized Validation
• Operates on a decentralized network of nodes (computers) that maintain a copy of the entire ledger.
• Rather than relying on a single centralized entity, all nodes work together to validate data.
• This consensus-driven validation process ensures that data is consistently checked across multiple
nodes, reducing the risk of fraudulent information.
16-04-2025 Dr. Malathi D., SCOPE–VIT 53
 Blockchain storage of Integrity Data
• Example: when a new block of data is added, nodes use a consensus mechanism—such as Proof of
Work (PoW) or Proof of Stake (PoS)—to agree on its validity. This system not only distributes trust
across the network but also prevents any single point of failure, making it harder for attackers to
manipulate data. Decentralized validation thus enhances data reliability, as data needs to be
independently confirmed by multiple nodes.

3) Smart Contracts
• Smart contracts are self-executing pieces of code stored on the blockchain. They automatically execute
predefined actions when specific conditions are met.
• For instance, a smart contract might release payment once a product delivery is confirmed.
• These contracts enable automated data processing and verification, which can be particularly useful in
applications requiring strict accuracy, such as financial services or IoT data validation.
• By eliminating the need for intermediaries, smart contracts foster transparency and trust in
transactions, as everyone can see and verify the terms and execution of each contract.
• Reduces human error and ensures that transactions are conducted exactly as agreed upon.

16-04-2025 Dr. Malathi D., SCOPE–VIT 54

 Blockchain storage of Integrity Data
4) Immutable Audit Trails
• Blockchain provides a built-in audit trail, allowing organizations to trace the history of data securely.
• Every transaction or data entry on the blockchain is timestamped and linked to previous records, creating an
unalterable chain of custody.
• Makes it possible to trace the source and progression of any dataset back to its origin.
• Example: In the healthcare industry, an immutable audit trail can be crucial for tracking patient data or
medication records, where accuracy is critical.
• Auditors and regulators can use this transparent trail to verify data legitimacy and identify any
inconsistencies, enhancing overall data integrity.
5) Data Attestation and Certification
• Blockchain can be used to certify data by allowing companies to attach a timestamp and cryptographic
signature to each dataset. This process is called data attestation.
• By certifying data on the blockchain, organizations can prove the authenticity and integrity of their data
without relying on external authorities.
• Example: Manufacturing company could use blockchain to certify quality control data for their products,
providing a record that shows each product met quality standards.
• This certification assures stakeholders, customers, and regulatory bodies that the data is accurate, reliable, and
in compliance with standards, simplifying regulatory processes.
16-04-2025 Dr. Malathi D., SCOPE–VIT 55
 Blockchain storage of Integrity Data
6) Supply Chain Transparency
• Blockchain enhances transparency and traceability within supply chains, which is especially valuable in
industries such as logistics, pharmaceuticals, and food production.
• Every transaction, shipment, or hand-off in the supply chain is recorded on the blockchain, creating a
comprehensive, verifiable record of the product’s journey.
• This transparency helps prevent fraud, such as counterfeiting, by enabling anyone to verify the
authenticity of a product's origin and movement through the supply chain.
• For instance, in the pharmaceutical industry, blockchain can be used to track drugs from manufacturer to
pharmacy, reducing the risk of counterfeit medications entering the market and ensuring that products
remain genuine and safe for consumers.

Summary
• Blockchain technology provides a combination of immutability, decentralization, automated smart
contracts, audit trails, data certification, and transparency. These features work together to protect data
integrity, prevent tampering, and build trust in the data's accuracy.
• By implementing blockchain, organizations across various industries can maintain secure, transparent, and
tamper-proof records, fostering better data security and confidence in data-driven decision-making.
16-04-2025 Dr. Malathi D., SCOPE–VIT 56
 Text Book and Reference Books
Text Book

• Dhillon, V., Metcalf, D., and Hooper, M, Blockchain enabled applications, 2017, 1st Edition, CA: Apress,
Berkeley.

Reference Books

• Diedrich, H., Ethereum: Blockchains, digital assets, smart contracts, decentralized autonomous
organizations, 2016, 1st Edition, Wildfire publishing, Sydney.
• Wattenhofer, R. P, Distributed Ledger Technology: The Science of the Blockchain (Inverted Forest
Publishing), 2017, 2nd Edition, Createspace Independent Pub, Scotts Valley, California, US.

16-04-2025 Dr. Malathi D., SCOPE–VIT 57