Microsoft Fabric monitoring
Monitor your Microsoft Fabric, capacity and cost.
Fabric capacity is hard to read. Smoothing and bursting hide what drives Capacity Unit consumption, and a single F SKU bill arrives with no breakdown by team. CloudMonitor folds your Fabric capacity into the same FinOps model as the rest of your Azure estate.
F2–F2048
Every Fabric SKU costed
CU usage
Tracked by workspace and item
Throttling
Surfaced before queries get rejected
Right-size
Scale, pause, or reserve, ranked by saving
The problem
Fabric capacity is a black box on the bill.
One capacity, no breakdown.
A Fabric F SKU bills as a single line. Finance cannot see which workspace, team, or item drove the cost.
Smoothing hides the spikes.
Bursting and smoothing keep Fabric fast, but they also mask which jobs are pushing a capacity toward throttling.
Reserve, scale, or pause? No signal.
Without utilization trends a capacity sits over-provisioned or gets throttled, and reservations go unbought.
What CloudMonitor monitors
Your Fabric capacity, read like the rest of your bill.
Capacity cost
F SKU spend, reserved versus pay-as-you-go, and idle capacity, costed in the same FOCUS model as the rest of your Azure bill.
CU consumption & throttling
Capacity Unit consumption over time, plus throttling and overage events, so a noisy workspace surfaces before it starts rejecting queries.
Right-sizing & reservations
Recommendations to scale an F SKU up or down, pause idle capacity, or buy a capacity reservation, ranked by dollar saving.
Workspace & item chargeback
Per-workspace and per-item cost across Lakehouse, Warehouse, pipeline, and semantic model, allocated to teams for showback and chargeback.
Optimization, not just charts
Right-sizing in your FinOps workflow.
CloudMonitor turns Fabric utilization into ranked actions: scale a capacity down, pause it when idle, or commit to a one-year reservation when usage is steady. Each lands in the same recommendation queue as your VM and storage savings, with the projected saving attached.
- Scale an F SKU up or down to match real CU usage
- Pause idle capacity and stop paying for it
- Reserve steady capacity for the committed-use discount
- Ranked by dollar impact, beside the rest of your Azure savings
Fabric capacity, ranked actions
Scale, pause, or reserve, each with a projected saving.
How it fits
Alongside the Fabric Capacity Metrics app.
Microsoft's Fabric Capacity Metrics app shows a capacity admin what is happening inside one capacity. CloudMonitor takes the same kind of signals and puts Fabric into your FinOps practice: allocated to cost groups, charged back to teams, and optimized beside every other line on your Azure bill. One model for the whole estate, not a separate tool per workload.
In the Data Platform scope
Fabric is the analytics tier. Synapse, Databricks, and Snowflake spend sit beside it in the same cost model.
Chargeback your teams expect
Per-workspace cost rolls into cost groups for showback and chargeback, with row-level security per team.
One estate, one view
Fabric capacity sits in the same reports as your VMs, storage, AI tokens, and licensing spend.
FAQ
Fabric monitoring questions
What does CloudMonitor monitor about Microsoft Fabric?
CloudMonitor reads your Microsoft Fabric the way it reads the rest of your Azure bill: F SKU capacity cost (reserved versus pay-as-you-go and idle capacity), Capacity Unit consumption with throttling and overage events, right-sizing and reservation recommendations, and per-workspace and per-item cost for chargeback.
How is this different from the Fabric Capacity Metrics app?
Microsoft's Fabric Capacity Metrics app shows a capacity admin what is happening inside one capacity. CloudMonitor takes the same kind of signals and puts Fabric into your FinOps practice: allocated to cost groups, charged back to teams, and optimized beside your VMs, storage, AI tokens, and licensing. It complements the Metrics app rather than replacing it.
Can CloudMonitor break Fabric capacity cost down by workspace and team?
Yes. A Fabric F SKU bills as one line, but CloudMonitor allocates it per workspace and per item across Lakehouse, Warehouse, pipeline, and semantic model, then rolls those into cost groups for showback and chargeback, with row-level security so each team sees only its own slice. See cost governance features.
Does CloudMonitor track CU consumption and throttling?
Yes. CloudMonitor tracks Capacity Unit consumption over time and surfaces throttling and overage events. Because Fabric smooths and bursts usage, a workspace can drift toward throttling unnoticed; CloudMonitor flags the capacity, and the workload pushing it, before queries start getting rejected.
Will CloudMonitor tell me when to scale, pause, or reserve a capacity?
Yes. When usage is low it recommends scaling an F SKU down or pausing idle capacity; when usage is steady it flags a one-year capacity reservation for the committed-use discount. Each recommendation is ranked by dollar saving in the same queue as your VM and storage savings. See optimization features.
Do I need to install anything in my Fabric capacity?
No. CloudMonitor connects through scoped, read-only access to your Azure billing and Fabric signals. There is nothing to deploy inside your Fabric capacity or your tenant, and CloudMonitor never writes to or changes your Fabric resources.
See your Fabric capacity in CloudMonitor.
Connect your Azure billing and your Fabric spend lands in the same reports the same day.
