Microsoft Fabric · Sydney
Microsoft Fabric Partner — Capacity Sizing & Power BI
Microsoft Fabric is Microsoft’s unified data and analytics platform.
combining Power BI, data engineering, data warehousing, real-time intelligence, and data science under one SaaS umbrella. Eagle360 helps Australian mid-market businesses with Fabric capacity sizing, workspace architecture, semantic model design, and migration from legacy Power BI and Synapse environments – including the rescue work when CapacityLimitExceeded errors hit production.
Capacity Sizing
The single most important Fabric decision
Microsoft Fabric is purchased as a single SKU called a “capacity” (F-SKU). The capacity determines everything: how many concurrent users, how big your semantic models can be, how fast your refreshes run, and whether you can use specific features like Direct Lake mode. Most Australian mid-market businesses get capacity sizing wrong on the first attempt — either over-buying out of fear or under-buying and hitting it within weeks.
F64 is the threshold at which Power BI consumers no longer need individual Pro licences — meaning at F64 and above, all internal users can view reports without per-user Power BI Pro fees. This is often the tipping point that makes Fabric cheaper than Power BI Premium per User for larger AU businesses.
Our Approach
A proven, four-step methodology for Fabric delivery
Eagle360’s Fabric methodology combines capacity sizing rigour with workspace architecture discipline and semantic model design experience built up over years of Power BI delivery.
Assess
Current state review, capacity sizing based on real concurrency, workspace architecture, and semantic model assessment.
Recommend
F-SKU recommendation, three-tier workspace plan, semantic model design, governance and DLP plan.
Implement
Lakehouse build, semantic model development, Power BI report delivery, source system integration.
Optimise
Capacity monitoring, semantic model tuning, DAX optimisation, and ongoing enhancement of reports and dashboards.
Ready to take your business to the next level?
Speak to our integration specialist.
Common Problem
Troubleshooting CapacityLimitExceeded errors
The most common Fabric production problem is CapacityLimitExceeded. This error means a workload (refresh, query, notebook, pipeline) exceeded the burst capacity of the current SKU. Symptoms include semantic model refreshes failing, report queries timing out, and pipelines being throttled or rejected.
Common causes (in order of frequency)
1. Trial environments running production workloads. Fabric trial capacities are F2-equivalent and intended for evaluation only. Production reports on a trial capacity will hit limits.
2. Wrong SKU for active user count. An F4 with 30 concurrent report users will choke. Active concurrency rather than named users is what matters.
3. Smoothing window exhaustion. Fabric smooths bursts over a 24-hour window. A poorly written refresh or DAX query that runs daily will eventually saturate the smoothing budget.
4. Bad DAX or model design. A single expensive measure can consume disproportionate capacity. Fixing the measure is cheaper than sizing up.
5. Direct Lake at the wrong SKU. Direct Lake requires F64+ and has specific behaviour at SKU boundaries.
Diagnostic workflow
Eagle360’s standard CapacityLimitExceeded diagnostic uses the Fabric Capacity Metrics App to identify the top consumers, then reviews semantic model design (DAX, model shape, refresh schedule) before recommending either a capacity increase or a workload optimisation. In most cases, a model fix saves more money than a capacity bump.
Workspace Architecture
Three-tier workspace pattern that scales
Fabric workspace architecture is where most projects accumulate debt fastest. A workspace structure that works for 5 reports becomes unmanageable at 50. Eagle360 follows a three-tier workspace pattern.
Tier 1 — Data foundation
Lakehouse and warehouse workspaces for ingested raw data, conformed dimensions, and gold-layer semantic models. Managed by the data engineering team.
Tier 3 — Self-service
Personal and team workspaces for ad-hoc analysis. Discovery happens here; promoted content moves to Tier 2.
Business Central + Fabric
The standard pattern for BC data in Fabric
For Eagle360’s Business Central clients, the most common Fabric scenario is centralised reporting across multiple BC environments (or BC plus other source systems). The standard pattern is:
1. Use Business Central’s native Microsoft Fabric integration to publish BC data to a Fabric lakehouse
2. Conform the BC data with other sources (CRM, e-commerce, finance subsystems) in the warehouse layers.
3. Build domain semantic models for Sales, Operations, and Finance with conformed dimensions
4. Deliver reports and Copilot-enabled dashboards from the semantic models
This pattern means Power BI reports refresh in minutes rather than hours, and historical reporting is not constrained by BC’s transaction-system performance characteristics.
Frequently Asked Questions
Everything you need to know about Microsoft Fabric in Australia
Can’t find the answer you’re looking for? Reach out to our team directly, and we’ll get back to you within 24 hours.
What is Microsoft Fabric, and how is it different from Power BI?
Microsoft Fabric is a unified SaaS data and analytics platform that combines Power BI, data engineering (Spark), data warehousing, real-time intelligence, data science, and data factory under one capacity. Power BI is one component of Fabric. The key differences are: Fabric provides a single OneLake storage layer that all engines share; Fabric is purchased as a single capacity (F-SKU) rather than per-user Power BI Premium; and Fabric includes engines beyond Power BI for data engineering and warehousing. For most AU mid-market businesses moving from Power BI Pro or Premium, Fabric F16–F32 is the typical starting point.
How do I choose the right Microsoft Fabric capacity SKU in Australia?
Capacity sizing depends on three factors: number of concurrent users, semantic model size, and data engineering workload. Indicative starting points for AU mid-market: F8 (~$1,680/month) for small teams with light Power BI; F16 (~$3,360/month) for typical mid-market with 40–100 concurrent users; F32 (~$6,720/month) for heavy Power BI plus regular pipelines; and F64 (~$13,400/month), at which free Power BI viewers are included. F64 is the threshold where Fabric often becomes cheaper than per-user Power BI Premium for AU businesses with 250+ report consumers.
What causes CapacityLimitExceeded errors in Microsoft Fabric?
CapacityLimitExceeded errors occur when a workload exceeds the burst capacity of the current Fabric SKU. The most common causes are production workloads running on trial capacity (Fabric trials are F2-equivalent and intended for evaluation only); the wrong SKU for the actual concurrent user count; the 24-hour smoothing window being exhausted by expensive recurring jobs; poorly designed DAX or semantic models consuming disproportionate capacity; and Direct Lake being used below the F64 threshold. Eagle360’s standard diagnostic uses the Fabric Capacity Metrics App to identify the top consumers and recommends either model optimisation or a capacity uplift.
Can I get Microsoft Fabric for a Business Central environment?
Yes. Microsoft Dynamics 365 Business Central has native integration with Microsoft Fabric. The standard pattern is to publish Business Central data to a Fabric lakehouse, conform it with other source systems in a warehouse layer, build domain semantic models for Sales, Operations, and Finance, and deliver Power BI reports and Copilot-enabled dashboards. This approach significantly improves reporting performance over running reports directly against Business Central.
How long does a Microsoft Fabric implementation take?
A Microsoft Fabric implementation for an Australian mid-market business typically takes 8–16 weeks for the first domain (usually Finance or Sales) and 4–8 weeks per additional domain after that. Greenfield implementations, including data ingestion, modelling, and report delivery, sit at the longer end. Migrations from existing Power BI Premium or Synapse environments are usually faster. Eagle360 delivers a structured discovery in 2 weeks that produces a fixed-price proposal.
Do I need Power BI Premium per user if I have Microsoft Fabric?
Power BI Premium per User (PPU) is not required for Fabric users at F64 capacity and above — all Power BI viewers (read-only consumers) can use Fabric without individual licences. Below F64, Power BI Pro or PPU is still required for users who view or share Power BI content. F64 is therefore the threshold at which Fabric often becomes cheaper than PPU for AU businesses with 250+ Power BI consumers.
What is Direct Lake mode, and when should I use it?
Direct Lake is a Fabric semantic model storage mode that reads directly from Parquet files in OneLake without importing into Power BI’s VertiPaq engine and without DirectQuery’s query passthrough. It combines the performance of import mode with the freshness of DirectQuery. Direct Lake requires F64+ capacity and is most valuable for large datasets where import mode refresh time is unacceptable, but DirectQuery query performance is too slow. For smaller AU mid-market scenarios, import mode is still the simplest and best choice.
Can Eagle360 fix a Fabric trial environment hitting capacity limits?
Yes. Trial Fabric environments are F2-equivalent and not intended for production. The fix is either to move workloads to a properly sized F-SKU capacity or to optimise the workloads so they fit within the constraints. Eagle360 has remediated several AU client environments where Fabric trial deployments were inadvertently used for production reporting. The standard remediation is a 2-week diagnostic plus a right-sizing exercise.
Right-size your Microsoft Fabric environment
Book a free 45-minute Fabric readiness review. We’ll assess your capacity needs, workspace plan, and governance approach — and tell you honestly whether you should size up, size down, or fix what you have.