I conducted recently an assessment for a client facing memory pressure in Power BI Premium. You know these pesky out of memory errors when refreshing a biggish dataset. They started with P1, moved to P2, and now are on P3 but still more memory is needed to satisfy the memory appetite of full refresh. The runtime memory footprint of the problematic semantic model with imported data is 45 GB and they’ve done their best to optimize it. This newsletter outlines a few strategies to tackle excessive memory consumption with large semantic models. Unfortunately, given the current state of Power BI boxed capacities, no option is perfect and at end a compromise will probably be needed somewhere between latency and performance.

Why I don’t like Premium licensing

Since its beginning, Power BI Pro per-user licensing (and later Premium Per User (PPU) licensing) has been very attractive. Many organizations with a limited number of report users flocked to Power BI to save cost. However, organizations with more BI consumers gravitated toward premium licensing where they could have unlimited number of report readers against a fixed monthly fee starting at listed price of $5,000/mo for P1. Sounds like a great deal, right?

I must admit that I detest the premium licensing model because it boxes into certain resource constraints, such as 8 backend cores and 25 GB RAM for P1. There are no custom configurations to let you balance between compute and memory needs. And while there is an auto-scale compute model, it’s very coarse and it applies only to processing cores. The memory constraints are especially problematic given that imported models are memory resident and require more than twice the memory for full refresh. From the outside, these memory constraints seem artificially low to force clients into perpetual upgrades. The new Fabric F capacities that supersede the P plans are even more expensive, justifying the price increase with the added flexibility to pause the capacity which is often impractical.

It looks to me that the premium licensing is pretty good deal for Microsoft. Outgrown 25 GB of RAM in P1? Time to shelve another 5K per month for 25 GB more even if you don’t need more compute power. Meanwhile, the price of 32GB of RAM is less than $100 and falling.

It will be great if at some point Power BI introduces custom capacities. Even better, how about auto-scaling where the capacity resources (both memory and CPU) scale up and down on demand within minutes, such as adding more memory during refresh and reducing the memory when the refresh is over?

Strategies to combat out-of-memory scenarios

So, what should you do if you are strapped for cash? Consider evaluating and adopting one or more of the following memory saving techniques, including:

• Switching to PPU licensing with a limited number of report users. PPU is equivalent of P3 and grants 100GB RAM per dataset.
• Optimizing aggressively the model storage when possible, such as removing high-cardinality columns
• Configuring aggressive incremental refresh policies with polling expressions
• Moving large fact tables to a separate semantic model (remember that the memory constraints are per dataset and not across all the datasets in the capacity)
• Implementing DirectQuery features, such as composite models and hybrid tables
• Switching to a hybrid architecture with on-prem semantic model(s) hosted in SQL Server Analysis Services where you can control the hardware configuration and you’re not charge for more memory.
• Lobbying Microsoft for much larger memory limits or to bring your own memory (good luck with that but it might be an option if you work for a large and important company)

Considering Direct Lake storage

If Fabric is in your future, one relatively new option to tackle out-of-memory scenarios that deserves to be evaluated and added to the list is semantic models configured for Direct Lake storage. Direct Lake on-demand loading should utilize memory much more efficiently for interactive operations, such as Power BI report execution. This is a bonus to the fact that data Direct Lake models don’t require refresh. Eliminating refresh could save tremendous amount of memory to start with, even if you apply advanced techniques such as incremental refresh or hybrid tables to models with imported data.

I did limited testing to compare performance of import and Direct Lake and posted detailed results in the “Fabric Direct Lake: Memory Utilization with Interactive Operations” blog.

I concluded that if Direct Lake is an option for you, it should be at the forefront of your efforts to combat out-of-memory errors with large datasets.

On the downside, more than likely you’ll have to implement ETL processes to synchronize your data warehouse to a Fabric lakehouse, unless your data is in Fabric to start with, or you use Fabric database mirroring for the currently supported data sources (Azure SQL DB, Cosmos, and Snowflake). I’m not counting the data synchronization time as a downside.

Teo Lachev
Prologika, LLC | Making Sense of Data

In a recent project, a client needed guidance on where to handle data quality and enrichment processes. Like cakes and ogres, a classic BI solution has layers. Starting upstream in the data pipeline and moving downstream are data sources, ODS (data staging and change data tracking), EDW (star schema) and semantic models.

In general, unless absolutely necessary, I’m against master data management (MDM) systems because of their complexity and cost. Instead, I believe that most data quality and enrichment tasks can be addressed efficiently without formal MDM. I’m also a big believer in tackling these issues as further upstream as possible, ideally in the data source.

Consider the following guidelines:

1. There are different types of exceptions and enrichment requirements, and they range in complexity and resolution domain. Each case must be triaged by the team to determine how to best handle it by examining the data pipeline upstream to downstream direction and traversing the layers.
2. As a best practice, exceptions and enrichments tasks should be dealt with as further upstream as possible because the further downstream the correction is made, the narrower its availability to consumers will be and more implementation effort might be required.
   1. Data Source – examples include wrong entries, such as wrong employe’s job start date or adding custom fields to implement categories requires for standard hierarchies. The team should always start with the source system in attempt to address data quality issues at the source.
   2. ODS – examples include corrections not possible in the upstream layers but necessary for other consumers besides EDW.
   3. EDW – examples include corrections applicable only to analytics, such as IsAttendance flag to flag which attendance events should be evaluated.
   4. Semantic layer – calculated columns for data enrichment best done in DAX and additional fields the end user needs in a composite model.
3. Shortcuts – When resolving the change upstream is time prohibitive, it should be permissible to handle the change further downstream while waiting for changes to an upstream system. For example, if fixing an issue in the data source is rather involved, a temporary fix can be done in ODS to prototype the change and expedite it to consumers. However, the team will remove that fix once the source system is enhanced.
4. When consumers prefer access to uncorrected data, a pair of fields could be introduced as upstream as possible, such as:
   1. EmployeeId – raw field
   2. EmployeeIdCorrected – corrected field. Joins to the related dimension table will be done on this field.