> ## Documentation Index
> Fetch the complete documentation index at: https://docs.thornode.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Monitor usage and limits

> Verify request activity, inspect per-method behavior, read live throughput, and compare traffic with published tier limits.

Use **Usage** to review application traffic and **Limits** to see the throughput and concurrency values for the selected access. [Plans and limits](/reference/plans-and-limits#understand-the-monthly-and-live-limits) explains monthly and live limit behavior.

## Verify traffic in Usage

<Steps>
  <Step title="Select the access and location">
    Open **Usage**, select the pass or rental, then choose the **Network** and
    **Location** used by your deployment.
  </Step>

  <Step title="Select a time range">
    Choose the range that contains the request you are investigating. **Usage overview** refreshes for that scope.
  </Step>

  <Step title="Confirm the request was recorded">
    Read **Total requests**, **Error rate**, **Requests over time**, and **Per-method analytics**. **View token summary** shows the access aggregate and representative key details; it is not a per-token activity table.

    After a test request, the request count or method activity should increase
    for the matching access, location, and range. Usage telemetry can arrive
    after the response, so use the application's status code for immediate
    debugging.
  </Step>
</Steps>

Use **Avg latency**, **p95 latency**, and **Latency over time** to compare behavior before and after a release. Scope both periods to the same access, location, and range before drawing a conclusion.

## Read published capacity in Limits

Open **Limits** and select the same access used by the application.

| Section             | Use it for                                                   |
| ------------------- | ------------------------------------------------------------ |
| **Live throughput** | Current request and transaction activity by location         |
| **Plan limits**     | RPC, WebSocket, and streaming limits for the selected access |
| **Consumption**     | Calls recorded during the selected reporting period          |
| **Calls by token**  | Comparing how traffic is distributed across location tokens  |

Limits can differ by access and product. Read the displayed values instead of encoding a tier table into your application. RPC caps apply per active location token where the table says so; **Calls by token** is the token-level usage view.

## Investigate throttling

When the application receives a rate-limit response or cannot open more concurrent work:

1. Confirm the access, network, location, and token used by the failing client.
2. Open **Usage** for the failure window and look for bursts, duplicate polling, retry loops, or unexpected methods.
3. Open **Limits** for the same access and compare traffic with the relevant published tier limit. For RPC, check the active location token under **Calls by token**.
4. Reduce concurrency, reuse connections, and close idle subscriptions.
5. Retry only transient failures with bounded exponential backoff and jitter.

Rotating a token does not increase capacity. Do not create extra tokens, locations, or reconnect loops to bypass the entitlement attached to an access.

## Diagnose dashboard signals

| Symptom                                | Verify                                                                    | Expected result                                                   | Action                                                                           |
| -------------------------------------- | ------------------------------------------------------------------------- | ----------------------------------------------------------------- | -------------------------------------------------------------------------------- |
| A request succeeded but Usage is empty | Access, location, time range, and ingestion delay                         | The request appears in the matching scope after telemetry arrives | Widen the range and compare with application logs                                |
| Usage is higher than expected          | **Per-method analytics**, access summary, and **Limits → Calls by token** | Methods and token rows match deployed services                    | Stop duplicate pollers or retry loops and rotate only if traffic is unauthorized |
| Live throughput says **No token**      | Selected access and location token                                        | The location has an active token under **Endpoints**              | Create the token or select the location that already has one                     |
| Client receives `429`                  | Relevant **Plan limits** row and **Calls by token**                       | Concurrency and rate remain below the published limit             | Bound work and back off transient retries                                        |
| Several products fail at once          | **Overview → Network status**                                             | No published incident explains the time window                    | If healthy, isolate the failing access and contact support with redacted details |

## Next steps

* [Configure, restrict, or rotate the location token](/dashboard/endpoints).
* [Design an RPC client for production limits](/products/thoredge-rpc).
* [Troubleshoot response and connectivity symptoms](/reference/troubleshooting).