You are here

NuoDB Technical Overview

To view this document as a PDF, click here.

Introduction | Technical Benefits | Use Cases | Architecture | Active-Active | Hybrid Cloud | Features | Specs 


THE NEED FOR A DISTRIBUTED HYBRID CLOUD DATABASE

Software development organizations understand that survival today means revisiting traditional approaches to application design, architecture, and infrastructure. As a result, organizations are increasingly offering services via elastically scalable, cloud applications in order to provide rapid, uninterrupted service and real-time user experiences.

Data is the lifeblood of any modern application. The credibility and success of your application depends on the user’s ability to reliably access and trust the data. However, one common misconception is that implementing high availability, elastic scale-out solutions requires signifcant data-related tradeoffs of traditional benefits, such as transactional consistency, data durability, rich SQL data access, and relational data models.

Instead of sacrificing and compromising data integrity, organizations should focus on implementing an elastic SQL database - an operational system of record database designed to provide ACID-compliant transactions and a standard SQL interface, as well as the elastic scale-out capabilities that modern applications require.

NuoDB is a full-SQL implementation that works very well both for on-premises deployments and in the cloud. It has an active-active deployment model with dynamic scalability and very good uptime characteristics.”

Tim Gage, Head of R & D, Alfa

 

INTRODUCING NuoDB

NuoDB’s elastic SQL database for modern applications combines the scale-out simplicity, elasticity, and continuous availability that cloud applications require with the transactional consistency and durability that databases of record demand. A true database of record designed to handle modern, dynamic operational workloads, NuoDB doesn’t force you to compromise traditional database benefits to achieve elastic cloud scalability.

With NuoDB, applications interact with the database as a single, logical, ACID-compliant, relational database via ANSI-standard SQL queries. However, NuoDB’s modular architecture allows organizations to elastically scale transaction and storage processes out and back in across data centers, deploy in any environment, and use popular orchestration tooling to manage and administer their database along with the rest of the application environment. 

logical diagram detailed

Figure 1. NuoDB’s modern, modular architecture allows organizations to elastically scale transaction and storage processes out and back in across data centers, deploy in any environment, and use popular orchestration tooling to manage and administer their database along with the rest of the application environment.

 

 

 

TECHNICAL BENEFITS

NuoDB was designed for modern applications, allowing organizations to:

elastic scaleElastically Scale to Meet Dynamic Throughput Demands

  • Easily and dynamically scale in or out to tune transaction performance for applications, on-demand
  • Handle unpredicted or predicted usage spikes and then return to normal load without needing to incur up-front hardware and license costs or recertify new hardware
  • Add transaction throughput without taking your application offline by eliminating the need for migration

caProvide Continuously Available, Active-Active Protection for Your Applications

  • Integrated, easy to operate capabilities which protect against process, system, and data center failures and require no additional disaster recovery testing
  • Automated and tunable data redundancy to eliminate failover time and provide uninterrupted service to the application
  • Perform rolling upgrades without outages
  • Achieve full active-active utilization of all available resources

support disaster recoveryDeploy Across Hybrid Cloud Environments to Support Disaster Recovery and Diverse Workloads

  • Deploy a single, logical database across multiple deployment platforms (physical, virtual, and containers) and environments (on-premises, multiple data centers, multiple clouds, hybrid cloud)
  • Transition from on-premises to the cloud on your own timeline using a single system of record
  • Build and support hybrid cloud architectures
  • Protect against cloud outages with a multi-cloud deployment strategy

diskObtain High Database Performance via Decoupled In-Memory Processing and Disk Durability

  • Serve up data to the application at in-memory speeds while maintaining and tuning data durability on disk-based storage
  • Eliminate the need for application sharding with dynamic and automated in-memory data management
  • Support hybrid workloads (e.g. HTAP) with in-memory processing and decoupled transactional and analytical processing

operationsOptimize Resourcing, Utilization, and Management with Straightforward Operations

  • Reduce operation costs with automatic replication and tunable redundancy
  • Straightforward to scale out and back in
  • Simplifed storage management eliminating the need for traditional fle system management tasks
  • Integrate with management, orchestration, and cloud ecosystems via RESTful APIs

sql databaseAvoid Risky Re-Writes with a Familiar Relational & SQL Application Model

  • Eliminate the need to re-write existing SQL application logic and lower migration costs with ANSI SQL compliance, including full DDL, DML, and ACID support
  • Data management logic is handled in NuoDB, freeing your developers to focus on building application features
  • Flexibility to quickly and easily perform ad-hoc queries, generate reports, and support complex relationships with adverse effects on database performance

 

 

 

USE CASES

NuoDB is designed to be a general system of record database for enterprise operational workloads, including online transaction processing (OLTP) and hybrid transactional/analytical (HTAP) processing workloads.

Well-suited as a technology for software development organizations that service a diverse range of industries, NuoDB is built for the modern data center, hybrid cloud environments, and for adaptability to efficiently handle a broad range of transactional workloads.

Popular use cases for NuoDB include software services within telecom, financial services, e-commerce, product lifecycle management, and healthcare management industries.

 

 

 

THE DURABLE DISTRIBUTED CACHE: AN ELASTIC SCALE-OUT ARCHITECTURE

NuoDB allows you to focus on developing applications, versus dealing with scale-out complexities. The database appears to the application as a single, logical, SQL database. However, under the hood, NuoDB features the Durable Distributed Cache, a modern architecture built for elastic scale out. In contrast to traditional monolithic databases, NuoDB uses separate services for transaction processing and storage management. This two layer, peer-to-peer architecture enables distributed processing that can be deployed across multiple data centers and is optimized for in-memory speeds, continuous availability, and elastic scale-out.

nuodb architecture

Figure 2. Unlike traditional monolithic databases, NuoDB separates the transaction processing layer and the storage layer, allowing each to scale out independently according to performance requirements.

The transaction processing layer consists of in-memory process nodes called Transaction Engines (TE). TEs allow NuoDB to maintain high, in-memory performance. As an application makes requests of NuoDB, TEs naturally build up in-memory caches with affinity for that application’s workload. Requests for data not currently in cache (cache misses) can be fulfilled from the memory caches of other TEs or from the storage management layer.

teTransaction Engines:

  • Handle requests from applications
  • Use in-memory processing to increase performance for applications
  • Coordinate transactions with other peer nodes in the environment
  • Can be added or removed without any interruption to the applications

The storage management layer consists of process nodes called Storage Managers (SM), which have both in-memory as well as on-disk storage components. The SM also caches data in-memory to speed up data retrieval from disk. SMs also provide on-disk data durability guarantees. Multiple SMs can be used to increase data redundancy.

smStorage Managers:

  • Ensure durability of data by writing it to disk
  • Manage data on disk
  • Serve up data to TEs from memory or disk
  • Maintain copies of data for redundancy
  • Can be added or removed without any interruption to application service

Without any need for application data management, and without any interruption to application service, NuoDB’s transaction processing and storage management services can elastically scale out (and back) simply by adding and removing TEs and SMs. 

summary network

Figure 3. NuoDB’s Durable Distributed Cache is composed of two layers of peer-to-peer, distributed process nodes that scale out independently, serve up data at in-memory speeds, and provide continuous availability for the application.

By presenting itself as a single logical database to the application, developers are freed from handling scale out complexity, continuous availability, and database management within the application code. Database operators can concentrate on scale-out, failure testing, maintenance and updates without worrying about adverse consequences to the application. As a result, both developers and operations can focus on truly maximizing performance of both the application and the database.

 

 

ACTIVE-ACTIVE, WITHIN AND ACROSS DATA CENTERS

Today, software organizations must meet customer expectations for always-on access to applications and uninterrupted service. Even the smallest amount of downtime can result in your users ditching your application in droves.

Traditional active-standby Disaster Recovery (DR) solutions generally do not offer continuous availability during a failure, and result in infrastructure inefficiencies as hardware sits idle or underutilized waiting for failures. Most DR solutions require some failover time to get the standby database ready for use. However, with an active-active deployment, you can eliminate this failover time and make full use of your previously idle infrastructure.

active data center

Figure 4. In traditional databases, if the active data center suffers a disaster, the standby data center must be brought online before the application becomes available again.

Active-active capabilities are a natural result of NuoDB’s peer-to-peer distributed, two-layer architectural design. While traditional databases require you to purchase expensive add-ons and NoSQL solutions offer high availability but not true ACID guarantees, NuoDB is always active, no matter how it’s deployed.

Composed of distributed process nodes, NuoDB can be deployed within a single data center, across multiple data centers, or even across hybrid cloud environments while maintaining ACID guarantees and serving up data at in-memory speeds.

However, NuoDB presents itself to the application as a single, logical SQL database. If one data center suffers a disaster, the rest of the database (in the other data center) can seamlessly take over the workload with no downtime.

multiple data centers

Figure 5. NuoDB can be deployed across multiple environments, providing active-active capabilities to all applications, simultaneously.

Applications with NuoDB don’t need to be aware of, or have to grapple with, typical active-active complexities such as coordination of independent active databases in different locations, conflict management or other typical active-active complexities. This frees up developers to focus on the important stuff - coding application features and functionality that really matter to the business.

 

 

 

 

HYBRID CLOUD DEPLOYMENT

Software organizations face major data hurdles when defining their cloud strategies, often believing they need to make significant compromises. However, your valuable data should be accessible to your application, anytime, and no matter where that application is located. With NuoDB, there is no need to make commonly believed sacrifices. Your data is not locked down in a data center or siloed in a cloud vendor environment. You do not need to struggle with syncing, accuracy, or data management issues.

NuoDB’s unique architecture allows your applications to access a single, logical, ACID-compliant database while deploying your data for access across:

  • Physical, virtual, and container platforms
  • On-premises, multiple data center, multiple cloud, and hybrid cloud environments

Here are some examples:

Migration to the Cloud:

migration to the cloud

migration to cloud

Figure 6. Keep your applications up and available while you migrate to the cloud.

Ideally, you’d keep your applications available and online while you migrate your data to the cloud. But, typical cloud migrations are complicated and require complex coordination, forcing your application offline while making the move. With NuoDB, you can provide continuous availability and up-to-date access to your data across both your traditional on-premises and new cloud environment as you undergo your migration. This gives you the flexibility to incrementally migrate your applications on your own timeframe and eventually shut down the on-premises side only after you are certain your cloud environment is functioning as required.

Availability During a Cloud Outage:

aws google cloud

Figure 7. Deploy NuoDB across multiple clouds to provide continuous availability in the case of a cloud outage.

Disaster recovery plans are an essential part of an organization’s cloud strategy. If your primary cloud environment suffers an outage, switching cloud providers or adding a new cloud provider while minimizing downtime and keeping data accurate can be a herculean effort. With NuoDB, it’s straightforward to deploy process nodes in the new cloud environment, add them to the existing NuoDB domain, and run an active-active database that’s designed to provide continuous availability, even in the event of a cloud provider outage.

Hybrid Cloud Application Architectures:

on prem aws

Figure 8. NuoDB allows you to deploy across hybrid cloud environments without needing ETL or data replication pipelines.

Software organizations that develop and run multiple applications often want to give those applications access to the same set of data. However, the typical approach to replicating and syncing data across your applications can cause huge data management headaches. If you are trying to do this across hybrid cloud environments, the issues become even more challenging to address. Avoid the complexities of managing multiple ETL or data replication pipelines by deploying NuoDB as your single system of record across all your applications.

With NuoDB, you can avoid data lock-in, serve data to your applications wherever it may be located, while preserving data and transactional integrity.

 

 

 

FEATURES AND FUNCTIONALITY

database functionalityCore Database Functionality

  • Operational system of record database with support for known and ad-hoc transactional workloads
  • Standard relational data model to support complex data relationships
  • Fully ACID compliant with tunable in-memory and disk-based durability
  • Multi-version concurrency control (MVCC) to maintain application performance as users scale
  • Modern database architecture with independent, peer-to-peer services for transaction processing and data durability

elastic scaleElastic Scalability and High Performance

  • Flexible configurations to optimize for performance, continuous availability, data redundancy, and other SLAs
  • Automated data locality and data affinity to provide in-memory performance
  • Elastically scale out and in transaction processing and storage management independently with no impact to application logic

caContinuous Availability and Recoverability

  • Built-in data and process redundancy for resiliency against network, hardware, and data center failures
  • Perform planned maintenance and rolling upgrades while the database stays online
  • Active-active capabilities across multiple data centers, Amazon Web Services availability zones, or hybrid environments without the need for additional add-on technologies
  • Peer-to-peer, non-hierarchical architecture that eliminates master/slave configurations or failover issues

sql databaseSQL Performance and Optimization

  • ANSI SQL compliant including support for DDL & DML, triggers, joins, ad hoc queries, stored procedures, and other advanced queries
  • Cost-based query performance optimization with full query explain plans
  • Optimization of common data access patterns using indexes
  • Support for both operational (OLTP) and fast analytical processing (HTAP) workloads

securitySecurity

  • Standard enterprise SQL database security features for user authentication and access controls
  • Network encryption across all database processes even across hybrid environments
  • LDAP integration for authentication 

operationsOperations

  • Integrated data redundancy and Disaster Recovery (DR) protection
  • Dynamic scale out and in that is transparent to application availability
  • Rolling upgrades to keep application online throughout the upgrade event
  • Deploy and manage as a single database across multiple data centers
  • Deployment on mixed platforms, including commodity hardware, VMs, or containers
  • Simple and flexible storage and data management
  • Easily integrate into existing operational platforms via RESTful APIs

 

 

 

SPECIFICATIONS

Deployment Hardware and Environments

  • Commodity servers
  • Virtual machines
  • Containers (Docker, Red Hat OpenShift Container Platform)
  • On-premises
  • Private cloud
  • Public cloud (e.g. Amazon Web Services, Microsoft Azure, Google Cloud Platform)

Operating Systems

  • Production:
    • Red Hat Enterprise Linux (RHEL)/CentOS 7.0 (x64) and above
    • RHEL/CentOS 6.5 (x64) and above
  • Test and Development:
    • Microsoft Windows 8 (x32) (client support only)
    • Microsoft Windows 8 (x64)
    • Microsoft Windows 10 (x64)
    • Microsoft Server 2012/R2 (x64)

Clients and Integration

  • JDBC / ODBC / .NET for client drivers
  • C/C++ native application library drivers
  • Standard RESTful APIs for monitoring and management integration

Networking

  • 10 GigE recommended

 

Consult with an Architect

Get expert advice on NuoDB's fit for your applications.

 


Not sure if NuoDB is right for you?
Schedule a session with one of our experts to find out.