# Nokku and NIS2

> How Nokku helps with the access, logging and incident duties NIS2 puts on server access.

This page maps what **NIS2** asks about access to your servers to what Nokku does. Use it when you write your own policies, or when an auditor asks how you control SSH.

NIS2 is the EU directive 2022/2555. It asks essential and important entities to manage the risks to their networks and systems, and Article 21(2) lists the measures. Nokku helps with several of them. It does not make you compliant on its own. Your policies, processes and people still do that.

Some of the features below may not be enabled on every instance.

## What Article 21(2) asks, and where Nokku helps

| Measure                         | How Nokku helps                                                                                                                                                                                                                                          |
| ------------------------------- | -------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- |
| (i) Access control              | Every login is a person on a named Linux account of a named server. Teams and SCIM keep it in line with your directory. Access can end by itself, or be asked for and approved. The access report shows who can reach what and flags grants nobody uses. |
| (j) Multi-factor authentication | Sign-in with passkeys or your identity provider, never a password. Single sign-on can be required for everyone.                                                                                                                                          |
| (h) Cryptography                | Short-lived certificates instead of SSH keys, keys kept in a TPM or on a security key where there is one, CA keys encrypted or protected by a KMS, recordings encrypted.                                                                                 |
| (b) Incident handling           | An audit log of every change and certificate, session recordings, revoking all access of a person at once, and streaming to your SIEM.                                                                                                                   |
| (g) Cyber hygiene               | No SSH keys to hand out, rotate or forget. Idle terminals can close by themselves.                                                                                                                                                                       |
| (d) Supply chain                | Run the core yourself. The CLI and the daemon are open source.                                                                                                                                                                                           |

## Evidence for an audit

- **Who has access:** export the [access report](/docs/security/access-review) regularly and keep it with the review notes. The audit log shows who exported it and when.
- **What they did:** the [audit log](/docs/security/audit-and-recordings) and the session recordings, linked to the person and the server.
- **That access ends:** grants that expire and approved requests are in the audit log, from the ask to the end.

## Reporting an incident

Article 23 asks for an early warning within 24 hours of a significant incident and a notification within 72 hours. Both go faster when the trail is complete:

- Stream the audit log to your SIEM, so it sits next to the rest of your logs and outlives the workspace's retention. See [Stream it to your SIEM](/docs/security/audit-and-recordings#stream-it-to-your-siem).
- Export what you need for the final report before retention removes it.
- **Deactivate** a compromised account. It is signed out everywhere, and servers with `nokkud` drop it and close its sessions at once, including what its teams gave it.
