# Rate limits & blocked IPs

View IP addresses currently throttled or blocked by threat protection, and release a block when needed.

A blocked IP address is one that the platform's threat-protection controls have throttled or shut out after too many failed attempts. The **Blocked IPs** view lists them live, and is most useful when a legitimate user has been locked out and you need to release their block.

You find it under **Audit Trail → Rate Limits** in the Admin Console.

> **Note:** Viewing the list and releasing a block require sufficient admin permissions. An administrator whose access is scoped to a limited set of resources may not see this view at all.

## How throttling and blocking work

The platform protects the login and signup surface against abuse by counting failed attempts and holding back the source once a threshold is crossed. Two kinds of limit drive the entries you see here:

- **Brute-force protection** — counts consecutive failed login attempts and blocks once the configured number is exceeded.
- **Per-IP rate limiting** — caps how many login attempts an IP may make per day and how many signup attempts it may make per minute.

These thresholds are configured under [Security → Threat protection](https://dts-identity.com/docs/security/threat-protection.md). Addresses inside a **trusted** [network zone](https://dts-identity.com/docs/security/conditional-access/network-zones.md) can be exempted from throttling, so a known office range is not held back by ordinary login churn.

### What trips each block, and what the client sees

| Limit | What trips it | What the throttled client experiences |
|---|---|---|
| **Consecutive login failures** | One account hit with repeated wrong passwords in a row — the classic single-account brute-force. | That account is locked out; further sign-in attempts are refused until the block clears or you release it. |
| **Login attempts per IP per day** | One source IP making a high volume of login attempts across many accounts in a day — credential-stuffing or scripted spray. | The IP is held back; its login attempts are refused for the rest of the daily window. |
| **Signup attempts per IP per minute** | One source IP firing many self-service sign-ups in a short burst — typically automated account creation. | The IP's sign-up attempts are refused until the per-minute window resets. |

The first is scoped to **an account**; the other two are scoped to **a source IP**. That distinction matters when you decide where to remediate: a single locked-out person is usually an account lockout, while a noisy IP affecting several accounts is throttling.

### Is it a real attacker or a false positive?

Before you release a block, check the context:

- **Look at the Target and Attempts.** A single account climbing steadily against the threshold is often a forgetful user or a stale saved password. One IP racing through *many* accounts is far more likely to be an attack.
- **Cross-check the [audit log](https://dts-identity.com/docs/monitoring/audit-logs.md).** Brute-force detection and rate-limit breaches are recorded there. Filter by the IP and time window to see whether the attempts target one account or spray many, and whether any succeeded.
- **Check the network zone.** Traffic from a shared corporate egress IP can trip per-IP rate limiting legitimately — consider adding that range to a [trusted zone](https://dts-identity.com/docs/security/conditional-access/network-zones.md) rather than repeatedly clearing it.

If it looks like a genuine attack, leave the block in place; it will continue to protect the surface. Only release a block you can attribute to a legitimate user.

## Reading the list

Each row shows:

| Column | Meaning |
|---|---|
| **Category** | The protection that created the entry. |
| **IP** | The source IP address being counted. IPv6 addresses are counted by their /64 prefix and shown as such (for example `2001:0db8:0001:0002::/64`). |
| **Type** | The specific limit (consecutive login failures, login attempts per day, signup attempts per minute). |
| **Target** | What the count applies to, where relevant. |
| **Attempts** | The current count against its threshold, with a progress bar. A row that has crossed its threshold is marked **Blocked**. |
| **Expires in** | When the count resets and the address is no longer held back. |

Blocks are temporary: each entry expires on its own once its window passes, and the address is then free again without any action on your part.

[Image: The Blocked IPs list with its brute-force and per-IP rate-limiting counters]

## Releasing a block

If a legitimate user is locked out and you do not want to wait for the block to expire:

1. Open **Audit Trail → Rate Limits**.
2. Find the row for the affected IP address.
3. Use the row's delete action to **unblock** it. The counter is cleared immediately and the address can attempt again.

> **Note:** Unblocking only clears the *current* count. If the underlying behaviour continues — for example a real brute-force attempt — the address will be throttled again. Investigate before clearing a block you did not expect.

## Related

- [Threat protection](https://dts-identity.com/docs/security/threat-protection.md) — where brute-force and throttling thresholds are configured.
- [Network zones](https://dts-identity.com/docs/security/conditional-access/network-zones.md) — define trusted ranges that bypass throttling.
- [Audit logs](https://dts-identity.com/docs/monitoring/audit-logs.md) — brute-force detection and rate-limit breaches are also recorded as audit events.
