Support for multidimensional models
Support for multidimensional models on Microsoft Azure Analysis Services
Currently focusing on remaining features to get us to GA with tabular. Will confirm timelines/availability for MD as soon as available.
Wes, can you please comment on this. I have customers waiting for MOLAP and Under Review for now nearly a year, is a bit much ! Is it happening or not ???
Please provide a stronger statement than 'as soon as available' about timelines for MD on Azure. It is feeling like never.
Wes, please reply as you are getting more and more votes on this vital feature, but it has gone completely silent.
Jon Burchel commented
@Azure AS PM Team Admin.
Tabular is now GA
Any update on MOLAP please ?
I get that the MS web security model is better suited to Tabular, but MOLAP is needed for complex financial calculations, and complex time based calculations e.g Insurance Analytics apps have two complex time dimensions Policy Start Date / Month / Year and Underwriting Month / Year which are interrelated. Tabular cannot handle these complexity or calculation wise.
Also from widening the adoption, MS with a hosted AS MOLAP option could kill the competition.
We are all in a catch 22 world at the moment with MS BI :
I) Everyone wants cloud
ii) Everyone wants performance and application flexibility
I can get I) with Azure Tabular AS and Power BI
I can get ii) With on premise MS OLAP AS and tools like our own EIB Analytics or Pivot Tables etc
I can't get both - Why ?
Bharath Kumar Kunjibettu commented
If Multidimensional is not going to be ready , please provide work around for features like .current member, Lead and Lag functions using DAX expression as re engineering is turning out to be challenging in Tabular Cube
We at Saasabi have an OLAP engine running as a PaaS platform on Azure. It is a parallel in memory OLAP supporting write-back, and the majority of the core functionality of SSAS. It supports MDX, XMLA, OLEDB, Odata and other. It can be deployed in multi tenancy environments or single tenancy. Sophisticated security model, supports multi authentication regimes. It has been in production for over 3 years. Has easy to extend ETL capabilities supporting combinations of API's as well as direct DB calls, extensible framework to integrate into cloud based workflows, ML, Cortana etc.
All benchmarks against SSAS on complex calcs and functions have shown multiple X performance advantages and more than double that again in terms of price performance.
We are a strong and committed MS partner and are committed to Azure. The Saasabi platform has been designed for complex planning application requirements. We have a client doing whole of government financial analytics consolidating data over 52 GL's. We have another partner extracting data from thousands of cloud accounting ledgers. If you need an OLAP engine on Azure, that is high performant, dramatically reducing your license costs (compared to VM based solutions) get in touch with Saasabi.
Agree with Michael & John, multidim is needed to support planning functionality.
B B commented
Guys, waiting for update / news on Multidimensional. Can Tabular & Multidimensional not developed in parallel instead of in tandem. Multidimensional users are being alienated from Azure AS.
Upvote AFPs comments
Christopher Raschke commented
Very much needed. Tabular does not support many to many relationships, writeback etc.
John Hagelin commented
I agree with Michael, Tabular doesn't offer any support to planning functionality, which is why Multidimensional AS would be a great step for the Azure stack in my opinion.
Michael Bjelvenmark commented
Of cource it's needed. To build an Corporate Performance Managament BI solution that's based on Financial Data Warehouse in Azure is support for multidimensional models needed. For example to get Planning functionality.
1) ROLAP against Azure DW - we want and need a fast data warehouse for large data sets, and we want our analysis tools to be in near real time, as the data warehouse loads, thus ROLAP against a data warehouse. Azure DW scales, and that is really important to the future of a DW
2) Multi-dimensional is better than tabular, has more features, is a better development tool
3) ROLAP reduces/eliminates the need to support large data sets in SSAS
4) ROLAP eliminates the need for dual data entry into both DW and SSAS