Replace your SSAS Cubes with Next-Generation OLAP on the Cloud and On-Premise Data Lakes

By August 29, 2019 November 13th, 2019 Uncategorized

Replace your SSAS cubes with next-generation OLAP on the cloud and on-premise data lakes
Are you struggling to get insights because the size of your data has outgrown the OLAP capabilities of SQL Server Analysis Services (SSAS)? Are you looking for a cubing solution that scales for massive data? In this blog, we will discuss the inherent challenges of SSAS OLAP, why it cannot scale for the modern data ecosystem, and how you can move your SSAS cubes to the cloud or an on-premise data lake effortlessly.

Key drivers for SSAS migration

SSAS OLAP offers a powerful way to aggregate data, and make it available to the business users for quick, ad-hoc analysis of enterprise data with BI tools such as Excel, Tableau, Power BI, etc. However, this widely used technology has some limitations when made to work on data at massive scale. Firstly, there is a limit to the size of data on which SSAS cubes can be built. Building a single cube holding all the data becomes impossible after a certain point. As a result, scalability becomes a significant issue. Secondly, when data volumes rise, the time required to process the cube increases significantly, leading to a delay in the availability of new data. With limited parallelism or support for distributed computing, it becomes difficult to fit SSAS OLAP in a modern data environment.

During our on-going interactions with our customers and prospects, we came across several enterprises that are looking to replace their SSAS OLAP cubes with alternative solutions. Some of the most common factors that are driving the shift are:

Platform Modernization
Enterprises migrate their existing legacy relational systems to modern cloud or on-premise big data environments so that they can leverage advanced storage technologies and utilize commodity hardware for computing. Once they modernize their data infrastructure, it makes sense to move their existing SSAS cubes and logic to the new platform.

Scalability
Scaling up and scaling out is a real challenge with SSAS when data volumes grow to terabytes and petabytes. As data volumes explode, enterprises look for a solution that can scale quickly to accommodate their growing data.

Performance
As concurrency increases, there is a considerable degradation in query performance with SSAS OLAP. This is a real bottleneck for large enterprises, and they seek a solution that can support high concurrency on massive data.

Cost Considerations
The need to leverage the cost benefits of commodity hardware on modern data platforms drives enterprises to explore alternative technologies. Instead of expensive, commercial, legacy relational systems, they look for solutions that can cash in on the advantages of cloud and on-premise data lakes.

Introducing Kyvos – the next-generation OLAP solution for massive data

Kyvos enables easy migration of your SSAS cubes to modern data platforms. It has been built with SSAS compatibility in mind and works on both on-premise as well as cloud platforms.

You should migrate from SSAS to Kyvos if you want to:

  • Build cubes on massive volumes of data without being limited by its size
  • Store all required information in your cubes and drill-down to deeper levels of granularity
  • Enable a large number of concurrent users to access the cubes without impacting performance
  • Leverage modern data platforms to scale quickly

Using its next-generation OLAP technology, Kyvos creates multi-dimensional OLAP cubes on massive data and stores them in a distributed manner across your data platform. The solution is highly scalable as it leverages modern data platforms to build, store, and query cubes.

The perfect solution for migrating your SSAS cubes

Kyvos is the best solution that helps you port your SSAS cubes to the cloud or an on-premise data lake, both because of its similarity to SSAS as well as the difference in the way it implements cubing, catering to modern data platforms.

  • High performance on massive data
    Both SSAS and Kyvos enable full pre-aggregation delivering high-performance, self-service analytics on enterprise data. However, the key differentiator for Kyvos is the scale of data that it can handle. Kyvos enables OLAP on trillions of rows of data, breaking the barriers of traditional OLAP capabilities. It offers the best of both worlds—the speed and interactivity of SSAS OLAP combined with the scalability and flexibility of modern data platforms. Though achieving OLAP at this scale is challenging, Kyvos uses its advanced technology to contain the combinatorial explosion that can happen while dealing with massive data.
  • Distributed architecture and parallelism
    Kyvos is designed for modern-day cloud and on-premise platforms. Its distributed architecture supports parallelism both for cube building as well as for querying. The Kyvos BI server cluster enables aggregation on data at a massive scale, and the query engine cluster delivers instant responses to queries. Since Kyvos uses the compute capacity of modern data platforms to build OLAP cubes, it can build cubes on massive data. This also helps in reducing cube building times as compared to SSAS where the cube building times go up considerably as the size of data increases. Additionally, Kyvos can support thousands of concurrent users without any impact on performance, something that is difficult to achieve in SSAS.
  • Cost-effective scaling
    Kyvos is designed to scale with your cloud or on-premise platform. Its scalable architecture makes it easy to add compute or storage capacity depending upon needs. If you want to accommodate more data, add more users, or reduce response times further, all you need to do is to add nodes to your cluster. However, in the case of SSAS, scaling becomes cost-prohibitive and limiting after a certain point. Buy a higher-end machine is more expensive than adding nodes to your cluster.
  • Standard access mechanisms
    Just like SSAS, Kyvos uses the MDX/XMLA connector. Besides, it also offers SQL connectivity. This enables your existing BI tools such as Excel, Tableau, MicroStrategy, Qlik, Power BI, and more, to connect to Kyvos. Additionally, Kyvos also provides its own visualization tool that offers native visualization capabilities with an intuitive drag-and-drop interface for self-service analysis.

Steps for Migration

Kyvos cube structure is similar to SSAS, making one-to-one migrations easy, without disrupting your existing analytical environment. This means that if you have an existing SSAS cube, you can port it to the cloud or on-premise big data platform relatively easily with Kyvos. All you have to do is to take the design of your SSAS cube and recreate it in Kyvos. Calculated measures can be dropped in as Kyvos understands and supports MDX. Besides this, some complex capabilities with SSAS cubes can also be ported to Kyvos.

The key steps for migration can be summarized as:

  1. Review your existing SSAS cube design
  2. Recreate the design in Kyvos
  3. Optimize design and build cubes
  4. Migrate BI logic
  5. Plug your dashboards to Kyvos

Kyvos installs directly on clusters located on the cloud or on-premise data lakes. It offers native support for all cloud platforms, including Amazon Web Services (AWS) and Microsoft Azure, and the latest releases of Cloudera, Hortonworks, MapR, and Apache Hadoop. There is no vendor lock-in, and you can work with any big data or cloud platform. Kyvos offers you the flexibility to use any BI tool and any data platform.

If you want to learn how you can migrate your SSAS cubes to a scalable cubing solution that delivers high performance on massive data, request a demo now.

Request Demo

Leave a Reply

one × four =