Quick wins for a faster PC:
Fix the driver behind crashes, sound loss and screen glitchesFind Drivers →Repair Windows errors before they cause bigger problemsFix Now →CRM API limits are not one universal number: they depend on the platform, API, environment, and the resource being protected. For high-volume automation, identify whether the constraint is request volume, execution time, concurrent calls, or an aggregate allocation; then pace work using the platform’s documented response and monitor both overall usage and individual requests.
Which CRM API limits apply?
Start by identifying the exact CRM product and API family. Dataverse service-protection throttling and Salesforce API allocations measure different things over different windows, so their published figures are not interchangeable capacity targets.
As an Amazon Associate I earn from qualifying purchases.
| Platform and limit | What is measured | Documented scope and window |
|---|---|---|
| Microsoft Dataverse service protection | Request count, combined execution time, and concurrent requests | Microsoft documents defaults of 6,000 requests and 1,200 seconds of combined execution time in a 300-second sliding window, plus 52 or more concurrent requests per web server. Microsoft says limits can vary between environments and change; each available web server enforces limits independently. These are defaults, not guaranteed capacity. Microsoft Dataverse API limits |
| Salesforce API allocation | Aggregate API calls | Salesforce describes an org-wide allocation over a 24-hour period. The applicable allocation depends on the org and API; consult that API’s documentation and the org’s usage information. Salesforce API Request Limits and Allocations |
| Salesforce long-running concurrency | Concurrent inbound calls lasting at least 20 seconds | Salesforce documents 25 calls for production and sandbox orgs and 5 for Developer Edition and Trial orgs. Salesforce states there is no concurrency limit for calls shorter than 20 seconds. Load and system issues can affect usable allocation. Salesforce API Request Limits and Allocations |
Dataverse: separate service protection from request entitlements
Dataverse service protection is intended to curb excessive API usage and protect shared availability. Its request, execution-time, and concurrency controls are distinct from Power Platform request entitlements, which concern accounting and allocations for requests through connectors. A 429 service-protection response is not itself a statement about an entitlement balance, and batching does not establish that entitlement accounting has been bypassed. See Microsoft’s API limits overview.
Do these 3 things before closing this tab:
1Repair Windows errors before they cause bigger problems2Scan for outdated or missing drivers - takes under a minute3Clear out junk files and repair common Windows errorsSalesforce: distinguish org allocation from long calls
Salesforce’s daily org-wide API allocation and its cap on long-running concurrent calls are separate constraints. A workload can encounter a concurrency issue even when its aggregate daily usage is a different concern; conversely, a short call is not covered by the documented 20-second concurrency limit but still contributes to API usage. The relevant API’s own documentation and the org’s current usage are necessary for capacity planning.
Why is an integration being throttled?
A limit-related failure can reflect a burst of requests, costly requests accumulating execution time, too many overlapping long calls, an aggregate API allocation, or platform resource protection. The same symptom—automation no longer making progress—does not identify which dimension is exhausted. Check the product, API family, response status and vendor error code before changing concurrency or retry behavior.
- Request-rate or service protection: A burst or sustained high rate can exceed Dataverse service-protection controls.
- Execution time: A set of individually valid requests can consume too much combined execution time within the applicable window.
- Concurrency: Overlapping or long-running calls may hit a concurrency constraint even if request volume is moderate.
- Aggregate allocation: The org may have used its API allocation for the relevant period.
- Entitlement accounting: Connector-related request entitlements are a separate question from Dataverse’s service-protection throttling.
How should clients handle 429 and REQUEST_LIMIT_EXCEEDED?
Dataverse Web API: honor Retry-After
When Dataverse service protection returns HTTP 429, the response includes a Retry-After value in seconds. A non-interactive client should wait that interval before sending another request; Microsoft says the duration reflects recent request demand. The SDK exposes a corresponding Retry-After value in fault details. Microsoft recommends beginning at a lower request rate, increasing gradually, and using the service’s feedback to guide pacing. Microsoft Dataverse API limits and Optimize performance for create and update operations.
Rank #2
Salesforce: interpret the vendor error in its API context
Salesforce documents REQUEST_LIMIT_EXCEEDED for API request-limit conditions, including excess long-running concurrent calls in its platform limits reference. Its REST error guidance says the code indicates that API request limits in the org have been exceeded in the documented context. The relevant Salesforce API determines the response behavior; do not assume every Salesforce limit produces an HTTP 429 or a Retry-After header. Consult Salesforce API Request Limits and Allocations and Status Codes and Error Responses.
The Tool Desk
Outbyte PC Repair FREERepair Windows errors before they cause bigger problemsFix Now →Outbyte Driver Updater FREEFix the driver behind crashes, sound loss and screen glitchesFind Drivers →Use a bounded recovery sequence
- Identify the source: Record the CRM, endpoint, and API family that returned the failure.
- Capture diagnostic details: Log timestamp, HTTP status, vendor error code, request or trace identifiers, elapsed time, and any retry metadata such as
Retry-After. - Classify the constraint: Determine whether evidence points to request rate, execution time, concurrency, aggregate allocation, or another resource-protection mechanism.
- Pause or reduce load: For Dataverse 429 responses, wait the specified interval. For other errors, follow the applicable API’s documented behavior rather than applying a generic retry rule.
- Resume safely: Use bounded retries and controlled queue processing. Before replaying writes, consider whether an earlier attempt may have succeeded; design deduplication or idempotency safeguards where duplicate writes would be harmful. This is integration design guidance, not a universal vendor guarantee.
- Verify recovery: Check that queued work drains and that error rates and latency return to acceptable levels before restoring higher concurrency.
How can teams monitor usage before automation fails?
Dynamics 365 and Dataverse
Microsoft’s environment-monitoring guidance directs administrators to Dataverse analytics for customer-engagement apps. For Finance and Operations, it identifies Dataverse and Lifecycle Services; that product family also has specific throttling-monitoring views for request usage and throttled requests. These are product-specific routes, not one universal Dynamics dashboard. Monitor API request limits.
Salesforce
Salesforce administrators can review usage in Setup’s System Overview, query the REST /limits resource, inspect the Sforce-Limit-Info response header, and configure API Usage Notifications. For request-level investigation, API Total Usage event logs can help attribute calls to a connected app or client and user; request IDs can help correlate event types. Access to event logs and history can depend on the org and product entitlement, so confirm availability in the target org. Salesforce API Request Limits and Allocations and Salesforce API limits and monitoring your API usage.
Build a useful operational view
Aggregate dashboards show whether overall consumption is approaching a limit; request-level records help explain which integration, client, user, or call pattern is responsible. Track both where the platform exposes the data, and alert before a launch or scheduled workload reaches its expected peak. Preserve enough context to correlate a burst in usage with a job run and its individual failed calls.
Independent reader supportYour contribution helps us test, update, and keep practical guides available for everyone.How can high-volume automation avoid avoidable throttling?
- Ramp up gradually. Establish a conservative starting rate, increase it in measured steps, and let platform feedback determine pacing rather than assuming a published default is safe for every tenant.
- Queue and pace work. Smooth large jobs instead of releasing every request at once, and reduce or pause workers when server feedback indicates pressure.
- Make requests efficient. Avoid unnecessary calls and expensive operations. Consider real-time integration patterns where they can replace periodic high-volume jobs, as Microsoft discusses in its service-protection guidance.
- Use batching selectively. Batching can reduce round trips in some designs, but very large batches can increase execution time, and concurrency and entitlement constraints remain relevant. It is not a universal throughput shortcut.
- Test recovery paths. Verify that workers respect retry signals, bound retries, preserve diagnostics, and avoid harmful duplicate writes before running a large production workload.
- Monitor before and during launches. Confirm that aggregate usage and request-level telemetry are available, then watch for rising latency, throttling, and queues that fail to drain.
What to compare when evaluating CRM API capacity
Before estimating how much automation an integration can sustain, compare the actual API families on the dimensions that govern their behavior. A single “requests per minute” figure may omit the constraint that matters.
Quick Recap
- What is counted? Requests, execution time, concurrent calls, or aggregate daily usage.
- What is the scope? User, application, organization, web server, or another resource boundary.
- What is the time window? A short sliding window and a 24-hour allocation describe different capacity constraints.
- What does failure return? Status, vendor error code, retry header, and any API-specific recovery contract.
- What can operators observe? Aggregate consumption, notifications, request-level events, and identifiers that connect calls to clients.
- Can the client recover safely? Consider pacing, queueing, replay, bounded retry behavior, and deduplication for writes.
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
