Teradata

ClearScape Analytics is an encapsulated set of functionality that is deployed alongside (and packaged as part of) other Vantage products, most notably VantageCloud Enterprise and VantageCloud Lake. This allows ClearScape Analytics to serve as the core analytics capability supporting any and every Vantage product. That said, it is worth noting that VantageCloud in particular provides an open analytics framework with which you can leverage your own or third-party analytics functionality rather than ClearScape Analytics.

In terms of concrete functionality, ClearScape Analytics supports a wide variety of languages for analytic purposes, most notably Python and R as well as Jupyter Notebooks, SageMaker, Azure Analytics, and more. Additionally, addon libraries for Python and R are available that allow you to generate SQL from your Python and R code and use that to run your queries. Parallel computation is available. It also offers graph capability, support for machine learning, and 150+ built-in analytical algorithms in support of advanced analytics and machine learning at scale. This includes time-series and temporal functions (over fifty of which are provided out of the box) that, along with comprehensive geo-spatial support, are combined in what Teradata calls “4D Analytics”. For machine learning in particular, you can either build models using ClearScape Analytics or import existing models from, for example, Dataiku, Spark, SageMaker, R, or Python. These models are converted to PMML (Predictive Model Markup Language), ONNX, MLEAP or an H20.AI model during the import process, and can then be executed in parallel just as your queries can.

In addition, ClearScape Analytics ModelOps is provided as a comprehensive means to manage your AI/ML models within Teradata Vantage. It aims to address the complexities associated with the deployment of AI/ML models, making it easier for you to leverage them effectively while ensuring that they continue to perform well over time. It offers capabilities in the areas of model lifecycle management, automated model deployment, model governance, and model monitoring. Specific capabilities of note include data drift and decay monitoring, update checking, and automated alerting mechanisms.

ClearScape Analytics provides connectivity plugins with Dataiku and H20.AI as well as API integration with AWS Sagemaker, Google Vertex AI, and AzureML.

Teradata has been the gold standard for data warehousing for several decades, offering extensive breadth and depth of analytical capability, and ClearScape Analytics both inherits and furthers that lineage. Moreover, while challengers have emerged, they have neither the breadth nor the depth that Teradata can offer. While machine learning support is increasingly common currency other vendors cannot typically compete with the capabilities offered by 4D Analytics, and rarely offer much in the way of ModelOps capabilities at all.

By bundling its analytics capabilities into ClearScape Analytics and allowing you to deploy it with any Vantage product, but especially with VantageCloud Enterprise or VantageCloud Lake, Teradata enables you to choose which kind of data architecture you want, what sort of querying you want to do, and what kind of analytics you want to use, essentially creating a modular analytics architecture. Moreover, ClearScape Analytics consistently provides benefits and drives positive business outcomes, regardless of your use of VantageCloud Enterprise or VantageCloud Lake.

The Bottom Line

ClearScape Analytics offers an exceptionally robust and well-proven set of analytics capabilities, including machine learning and ModelOps. In short, there is a lot to recommend it.

Teradata supports a wide variety of languages for analytic purposes, most notably Python and R, as well as Jupyter Notebooks. It also offers an AnalyticOps Accelerator, a collection of best practices, proven design patterns, and tried-and-tested code derived from successful Teradata projects, that is designed to make it much easier for you to implement an AnalyticsOps framework.

The platform offers graph capability, support for machine learning, and access to “Advanced SQL”, Teradata’s extension to SQL in support of advanced analytics and machine learning at scale. This includes time-series and temporal functions that, along with comprehensive geo-spatial support, are combined in what Teradata calls “4D Analytics”.

Teradata also provides a wide range of industry and analytic models.

All of the capabilities described above are delivered in the cloud just as they are on-premises, allowing Teradata Vantage to be used as a highly effective cloud data analytics platform. It can be deployed across a number of clouds, including the “big three” of AWS, Microsoft Azure, and Google Cloud, as well as VMware, Teradata Cloud, and purpose-built on-premises infrastructure.

Multi-cloud, hybrid cloud, and hybrid multi-cloud are all supported, and in terms of the cloud the product is generally delivered as a service. Any number of instances can be viewed, managed and monitored using the same web interface, which is particularly important for multi-cloud. In addition, Teradata Vantage is a highly portable solution (including its licensing – see below), largely owing to the fact that it uses the same software across all platforms. Teradata also provides data migration tools and best practices, and combined with the company’s experience with cloud migrations, moving an instance from one cloud to another should be a relatively painless process.

Fig 01 - Teradata Vantage ‘Big 3’ cloud integrations

What’s more, the platform’s relationship with these clouds reaches significantly beyond just deployment.

In fact, Teradata Vantage is closely aligned with all three major clouds, and boasts integration with a wide variety of Microsoft, Amazon and Google cloud services, as shown in Figure 1. Moreover, Teradata Vantage offers a software architecture (see Figure 2) that is well suited to the cloud, which helps to take advantage of the aforementioned cloud services provided by Azure, AWS and GCP.

Fig 02 - Teradata Vantage cloud architecture

Pricing models for Teradata Vantage in the cloud are flexible, consisting of both capacity-based pricing (‘blended’) and pay-as-you-go usage-based pricing (‘consumption’). The latter, in particular, provides automatic elasticity, and you only pay for successful queries and loads. It also boasts highly trackable usage statistics (which makes for predictable pricing), and departmental chargeback, among other things. In short, you are able to choose whichever pricing model suits your needs, most likely in terms of high or low (or unknown) utilisation. Also notable is that all Teradata Vantage cloud deployments are single-tenant, which has benefits for both performance and stability, and that Teradata’s pricing models and licenses are portable across cloud (and even on-premises) environments.

Teradata Vantage also offers separation of storage from compute, which helps support these pricing models (particularly consumption pricing). That said, it is not strictly required. Elastic scaling, dynamic resource allocation, software performance optimisation (via indexing and determining least-cost execution methods), and workload management are all available, and all contribute to performance in one way or another. Dynamic resource allocation also addresses data replication and data drift, as well as query prioritisation more complex than ‘first come, first serve’.

Teradata has been the gold standard for data warehousing for several decades. While new challengers have emerged over the last few years, they have neither the breadth nor the depth that Teradata can offer. While machine learning support is increasingly common currency other vendors cannot typically compete with the capabilities offered by 4D Analytics. This will be particularly true within IoT (Internet of Things) environments but is by no means limited to those use cases.

As a cloud solution, Teradata admixes its signature, widely-regarded analytics with the benefits of the cloud. This most notably includes increased flexibility, scalability, and ease of use. Starting with a modest Teradata cloud deployment, paying only for what you use and scaling up precisely as much as you need to also removes much of the onboarding difficulties that often come with a solution as broad and fully-featured as Teradata, while at the same time reducing risk and hastening time to value.

The Bottom Line

Teradata Vantage is well-known for its performance, scalability, high availability and reliability. These qualities are retained, and in some cases enhanced, in cloud deployments. In short, with its strong history, proven use cases and exceptional capabilities, Teradata Vantage should be considered for any cloud and/or on-premises analytic database requirements.

VantageCloud Lake is designed to let you run analytics on data in the object store and other systems while leveraging the push down capabilities of Teradata’s data fabric. This provides the ability to move the query to the data rather than incurring the costs and complexity of moving the data itself. It provides a self-service, web-based console for doing just that, as well as an open, connected analytics framework, enabling you to either leverage analytics capabilities provided by ClearScape Analytics or to bring your own. Third-party API support and integrations are also featured, and collaboration is supported via data sharing and the Teradata store.

VantageCloud Lake uses a new, consumption-based pricing model that offers optimised, unit-based pricing. It also offers a variety of spend management features. For instance, it provides automatic elasticity with incremental, intelligent scaling, meaning that you only scale up when your currently allocated resources reach their limits, and you only pay for the amount of data actually accessed or stored rather than just provisioned. The product boasts highly trackable and comprehensive usage reporting, which makes for predictable pricing and holistic financial visibility, and you can put guardrail policies in place on particular workloads to prevent accidental overspending. In addition, VantageCloud Lake separates storage from compute, which helps support this pricing model by enabling granular chargebacks to compute cluster owners. Multi-cluster compute is available.

The product also features several built-in resiliency features, such as backup-as-a-service, and rolling upgrades delivered via CI/CD pipeline. High availability is provided, and is enhanced via the session manager, which will hold user sessions and queries in place even if the application they are using goes down.

At present, VantageCloud Lake is available on AWS and Azure, with Google Cloud integration planned for 2024.

Teradata has been the gold standard for data warehousing for several decades, offering extensive breadth and depth of capability, but with VantageCloud Lake it has have taken a step forward to bring its well-developed analytics capabilities to wherever your data lives.

This provides considerable benefits to ad hoc and exploratory querying, which by their nature require low up-front costs in terms of time and effort expended by the user in order to be efficient. Essentially, having to move your data (and quite likely a lot of it to, since the whole point of exploratory testing is that you do not know precisely what you are looking for) to a data warehouse in order to perform a one-off query on it is a lot of wasted effort, and in such a scenario we would imagine that exploratory and ad hoc queries just do not happen, or if they do are a laborious process that everyone would rather avoid doing. VantageCloud Lake provides alternatives by allowing you to move your data to low-cost storage (such as the object store) and by enabling you to move your query rather than your data.

Past that, VantageCloud Lake possess many of the same strengths and capabilities as Vantage in the Cloud before it, but is now deployed in a much more cloud-focused architecture. As such, it benefits from Teradata’s considerable analytics capabilities as well as enjoying all of the advantages of the cloud, including flexibility, scalability, and ease of use.

The Bottom Line

Teradata VantageCloud Lake is by no means a complete replacement for more traditional enterprise analytics (at least, not yet), as seen in, say, VantageCloud Enterprise, but it is certainly a compelling complement to it. At the same time, it – and VantageCloud in general – has retained many of the qualities that have made Teradata Vantage so well-regarded in terms of performance, scalability, high availability, reliability, and so on. In short, it is well worth considering for inclusion in your analytics suite.

Fig 02 - Integrate data without moving or duplicating it

Teradata’s QueryGrid feature allows a virtual window into data, wherever it may be sourced, and presents this data to end users. Semantic mapping enables data to be presented in a user-friendly format, eliminating the need for users to be familiar with the underlying physical structures. This layer goes as far as handling deduplication of data and consolidation of data sources where appropriate.

Teradata has a partnership with Microsoft, and this includes working with the Microsoft Data Fabric product set. Teradata’s AI Unlimited, currently in private preview, will work with the OneLake logical data lake within the Microsoft Fabric environment. However, Teradata’s Data Fabric offering can expand the reach of Microsoft Fabric to non-Fabric data sources; it is offered as an option alongside, aimed at customers who have already committed to the Microsoft product.

Teradata Vantage allows customers to construct an analytics environment without necessarily copying data from sources into a separate data warehouse: this is the goal of a data fabric. In practice, this approach has some limitations in specific instances where there are performance SLAs, where it could be better if the data is localized and tuned in the processing engine of choice. In reality, some data will be more naturally materialised and re-used in queries. However, Teradata Vantage makes this approach transparent to the end user and allows each customer to be as virtual or otherwise as they wish to be in the way they set things up.

The recent interest in data fabric has caused a lot of vendors to frantically re-label what they had before as data fabric-friendly. However, some things are much harder than others to build. Producing a knowledge graph from a catalog is one thing, but building a proper distributed query engine with an effective optimiser is a much more complex task. Teradata is fortunate in that they have long had the foundations for this, based on their heritage in high-end data warehousing.

They have a well-proven semantic layer, a catalog, a knowledge graph, good source data integration and above all a proper distributed query optimiser, so the key elements of this architecture are all in place. Many data fabric solutions require stitching together several elements from different vendors or open-source solutions. Teradata is about as close as you will find to a complete offering in the data fabric world.

The bottom line

Teradata has built on its long heritage in highly scalable data warehousing and have extended this deftly into the data fabric world. Moreover, they have many customers actually using this approach in practice. For a proven data fabric solution versus something shown via PowerPoint or a canned demo, Teradata should be high on your list to consider.