Fabric Trial - Insufficient evidence to purchase - What next?

[Lobstersrnice]

Hello,

We have been trialing Fabric in the business for a lengthy period but have only recently taken on a Data Engineer and attempted to create a proof of concept for Fabric to replace many of our current reporting solutions.

The trial has been disapointing both in experiencing down time, pipeline failures and huge CU usage when performing fairly standard Spark Notebooks within a custom environment. We have queried and spoken to MindTree several times but between a lack of knowledge on our side, confussing usage and pricing data and a lack of clarity from the documentation, we have not proved to the busienss that Fabric is the solution.

We have been told there is no chance of extanding the Trial and therfore are likley to look at other options. This is not desirebale given that we are a microsoft house so reaching out for anyones suggestions as to how to move forward, has anyone had extensions beyond 60 days or more than 5 user trails? Is anyone from Micrsoft able to help us look at out proof of concept and tell us why it is so inefficient?

[Greg_Deckler · Accepted solution]

@Lobstersrnice Pretty sure you are not the only one in this boat. The usage is crazy complicated to decipher. The high CU usage isn't necessarily a problem with the bursting and smoothing but it becomes extremely complicated to sort of tell when you might run into problems and thus you are likely to over license your workloads. 

I was down to like 10 days left on my Fabric trial ( which has been running for about a year and a half ) and it was prompting me if I wanted to extend my trial but now suddenly I am back to 28 days of trial left. What you might try is to spin up a new Power BI tenant and then spin up a trial of Fabric.