Cosmos DB RU Cost and Throttling Monitor

Overview

This solution guards Azure Cosmos DB performance and cost. On a 30-minute schedule it queries Azure Log Analytics (KQL over Cosmos diagnostic logs) for RU consumption and 429 throttling rates per container, flags containers that breach a configurable throttle threshold, and posts a formatted alert with a scaling recommendation to an engineering Microsoft Teams channel.

Why it matters: RU exhaustion causes throttling (429s) and outages, while over-provisioning wastes money. Continuous monitoring keeps Cosmos right-sized and healthy, and surfaces hot containers before they impact users.

Status: Built as a FlowLibs reference demo. Ships Off. Going live requires only authorizing the Azure Log Analytics connection and setting the environment variable values - no logic changes.

Use Case

A team running Azure Cosmos DB wants early warning on throttling and cost. Rather than manually watching the Azure portal metrics blade, this flow runs on a schedule, evaluates the last 30 minutes of data-plane telemetry, and only pings the team when a container actually exceeds the alert threshold - turning reactive firefighting into proactive capacity management.

Flow Architecture

Environment Variables

Connectors & Connections

Connector	API name	Actions used
Azure Log Analytics	shared_azureloganalytics	RunOmsQuery
Microsoft Teams	shared_teams	PostMessageToConversation

Note — All connections are referenced as solution connection references; the flow is portable between environments as long as a connection is mapped at import time.

Customization Guide

Almost every realistic variant of this flow can be implemented by changing environment variable values. A few cases require small edits inside the flow definition — those are called out explicitly below.

Tune the threshold

Raise or lower flowlibs_ThrottleThreshold (the value is injected into the KQL where ThrottleRatePct >= clause at runtime).

Change cadence

Edit the Recurrence interval (e.g. 15 min for tighter SLAs).

Hot-partition detection

Add a second KQL block using CDBPartitionKeyRUConsumption to identify skewed partition keys, and surface the top partitions in the alert.

Autoscale toggle

Extend the True branch with an Azure Resource Manager Resources_CreateOrUpdateById call to enable autoscale on the flagged container automatically.

Cost trend

Persist each run's RU totals to a Dataverse table for week-over-week spend reporting.

Key Expressions

The flow is intentionally light on Power Fx / WDL gymnastics — the heaviest expressions are the branch-name concatenation and the approval outcome check. They are listed below in the order they appear in the flow.

@concat('AzureDiagnostics | where ResourceProvider == "MICROSOFT.DOCUMENTDB" | where Category == "DataPlaneRequests" | where TimeGenerated > ago(30m) | extend isThrottled = iff(statusCode_s == "429", 1, 0) | summarize TotalRequests = count(), ThrottledRequests = sum(isThrottled), TotalRU = round(sum(todouble(requestCharge_s)), 1) by AccountName = databaseAccountName_s, CollectionName = collectionName_s | extend ThrottleRatePct = round(100.0 * ThrottledRequests / TotalRequests, 2) | where ThrottleRatePct >= ', variables('varThrottleThreshold'), ' | project ... | order by ThrottleRatePct desc')

@greater(length(coalesce(body('Query_Cosmos_Metrics')?['value'], createArray())), 0)

@replace(replace(replace(string(body('Select_Flagged_Rows')), '"},{"html":"', ''), '[{"html":"', ''), '"}]', '')

Customize & download