fix: the web service waits for the PostgreSQL driver instead of failing at start.

feature: bootstrap diagnostics on port 17081 during the first setup.
feature: the web service creates its key and database itself and shows its starting phase.
2026-10-24 10:43:51 +02:00 · 2026-10-24 10:07:22 +02:00 · 2026-10-24 09:07:28 +02:00 · 2026-10-24 08:55:25 +02:00 · 2026-10-24 04:19:01 +02:00 · 2026-10-24 03:00:35 +02:00
23 changed files with 1393 additions and 247 deletions
@@ -52,27 +52,31 @@ configuration. It runs as root, because `swanctl` and `wg` need it, with `NoNewP
 `ProtectSystem=strict`, `ProtectHome` and `PrivateTmp`. The Tailwind style comes from a CDN, so a
 browser on a network without internet shows the unstyled page, and `curl` is unaffected.

-## ADR-3: /config is authenticated with a short-lived JWT signed with a shared secret
+## ADR-3: The protected paths take JWTs that only the router signs

 Status: Accepted

-Context: A client such as the Router Manager needs the configuration and the resolved state of
-the router, which `/` and `/status` must not show. The service listens on plain HTTP on the
-internal network, and its configuration file is readable by root only. The platform that deploys
-the router generates one secret and gives it to the router's configuration and to the client.
+Context: Clients such as the Router Manager, and people with `curl`, need the configuration,
+the users and the groups of the router, which `/` and `/status` must not show. Both people and
+programs sign in with a user of the directory. A secret shared with the platform that deploys
+the router would let whoever holds it, and every copy of the deployment's state, make tokens.
+The service listens on plain HTTP on the internal network.

-Decision: `/config` needs a JWT in an `Authorization: Bearer` header. The service verifies HS256
-tokens with `hmac` and `hashlib`, with the secret `web.jwt_secret`, and accepts no other
-algorithm, because a shared secret is what both sides have; RS256 or ES256 would need a key pair
-and its distribution. The token must carry `exp`, not more than ten minutes ahead, so that the client makes a fresh token for
-each use and a leaked one expires. Without a secret the page is not served. The configuration
-shown has no secret options (ADR-8).
+Decision: `/config`, `/users` and `/groups` need a JWT in an `Authorization: Bearer` header,
+HS256, verified with `hmac` and `hashlib`; no other algorithm is accepted. The key is
+`/etc/vpn-router/web.key`, random, made by the web service on its first run and readable by root
+only; it is never configured or output. The router itself issues
+the tokens: `POST /auth` checks a user's password and TOTP code with a bind (ADR-14) and answers
+an access token naming the user in `sub`, valid for 5 minutes, and a refresh token valid for 8
+hours (ADR-15); `vpn-router-token` makes tokens with the `admin` claim (ADR-13). A token without
+`admin` must expire within ten minutes, so a leaked one is of little use. Without the key the
+paths and `/auth` are not served. The configuration shown has no secret options (ADR-8).

-Consequences: The client and the router share one secret, and whoever holds it can make tokens,
-so it is long and random; it is checked to have at least 32 characters. The secret is only in the
-configuration file, readable by root only. The token and the answer cross the network
-unencrypted, so the page is for a trusted internal network; HTTPS would need a certificate that
-the package does not have.
+Consequences: Nothing outside the router can make a token, so a token is either issued to a
+user who signed in or made by root on the router. A client keeps the refresh token to stay
+signed in. A token of `vpn-router-token` can be revoked only by replacing the key, which ends
+every token. Tokens, passwords and answers cross the network unencrypted, so the API is for a
+trusted internal network; HTTPS would need a certificate that the package does not have.

 ## ADR-4: Users are kept in a local OpenLDAP directory

@@ -144,8 +148,8 @@ the CA's state, and `/etc/swanctl/x509crl` for the published CRL.

 Consequences: The rules exist once. A request starts no process, and a password never leaves
 the process that received it. The web service imports `python3-ldap` and the package's
-configuration module. The API is protected by the token of `/config`, so whoever has the secret
-can manage users.
+configuration module. The API is protected by the tokens of ADR-3, so only members of `admins`
+and the holders of an admin token manage users.

 ## ADR-7: The web service starts on its own and reads its inputs at each request

@@ -161,14 +165,14 @@ thread works out, every 30 seconds, the internal address from `vpn-router.json`
 too, and `web.port` and `web.enabled` from the same file. It runs the HTTP server in a thread of
 its own at that address and port, stops it and starts a new one when they change, and runs none
 while the service is disabled, the address is not known or it cannot be bound. Each request
-reads the token secret from `vpn-router.json` and the features and resolved facts from
+reads the signing key and `vpn-router.json`, and the features and resolved facts from
 `web.json`, which setup writes when it applies the configuration, so the service shows what setup
 applied; while that file is missing the router shows no features. The unit is not
 ordered after `vpn-router-setup`, and setup does not restart it.

 Consequences: The status page answers as soon as the internal address exists, whatever state
 setup is in, and follows a changed address, port or `web.enabled` within 30 seconds without a
-restart. A change of the secret applies to the next request. The interfaces are read every 30
+restart. A new signing key applies to the next request. The interfaces are read every 30
 seconds, which costs a few `ip` commands.

 ## ADR-8: /config leaves secret options out
@@ -179,8 +183,8 @@ Context: A secret has no use outside the router, so a client never needs it, and
 that a secret is set tells a reader more than it needs.

 Decision: The configuration that `vpn-router-setup` writes for the web service has no secret
-options at all. They are recognised by name: `psk`, and options with `secret`, `password`,
-`token` or a `_key` ending in the name.
+options at all. They are recognised by name: `psk`, `database`, whose connection string may hold
+a password, and options with `secret`, `password`, `token` or a `_key` ending in the name.

 Consequences: No secret reaches the web service's state file or any answer of the API. A client
 cannot tell whether a secret is set. A new secret option needs a name that matches.
@@ -190,26 +194,28 @@ cannot tell whether a secret is set. A new secret option needs a name that match
 Status: Accepted

 Context: The router needs a web application to manage it, which may run on the router or on
-another Linux system. The router's web service needs a token signed with `web.jwt_secret` for
-`/config` and `/users` and sends no CORS headers, so a browser cannot call it with the secret
-kept away from the page.
+another Linux system. The router's web service issues tokens through `/auth` (ADR-3) and sends
+no CORS headers, so a browser cannot call it directly, and tokens kept in the page could be read
+by its scripts.

 Decision: `vpn-router-manager` is a Debian package of its own, built in the repository
 `linux-cloud-router-manager`. A React and MUI application in the style of manage-ui is built
 with Vite when the package is built and shipped as static files. A Python server, using the
-Ubuntu packages `python3-jwt` and `python3-requests`, serves them and passes every `/api/<path>`
-call to the router's `/<path>`, signing a short-lived token with the router's secret; it holds
-no logic of the router's own. It reads its settings from `/etc/vpn-router-manager/manager.json`
-at each request; on the router the settings not given come from `vpn-router.json`, and it
-listens on the internal address like the status web service, with the same discovery thread.
-`vpn-router-setup` opens the manager's port on the internal interface while the manager is
-installed, so the firewall stays in one place. The manager asks for no authentication.
+Ubuntu packages `python3-requests` and `python3-psycopg`, serves them, signs users in through the
+router's `/auth`, keeps their tokens in a server-side session in PostgreSQL behind an
+`HttpOnly` cookie, and passes every other `/api/<path>` call to the router's `/<path>` with the
+session's access token, refreshing it as needed; it holds no logic of the router's own and no
+key of it. It reads its settings from `/etc/vpn-router-manager/manager.json` at each request; on
+the router the settings not given come from `vpn-router.json`, and it listens on the internal
+address like the status web service, with the same discovery thread. `vpn-router-setup` opens
+the manager's port on the internal interface while the manager is installed, so the firewall
+stays in one place.

 Consequences: The router needs no Node.js; the package needs only `python3` and packages of the
-Ubuntu archive. The secret never reaches the browser. The application checks the shape of every
-answer of the router. Anyone who reaches the manager's port manages the router, so on another
-system it listens on `127.0.0.1` unless told otherwise. The router's firewall depends on whether
-the manager is installed, and a changed manager port needs `vpn-router-setup` to run again.
+Ubuntu archive. Tokens never reach the browser, and sessions survive a restart of the manager.
+The router decides what a user may do; the manager only signs in. The application checks the
+shape of every answer of the router. The router's firewall depends on whether the manager is
+installed, and a changed manager port needs `vpn-router-setup` to run again.

 ## ADR-10: Users and their certificates do not depend on a P2S system being enabled

@@ -247,3 +253,145 @@ serves them under `/groups`. Group names follow the rules of user names.
 Consequences: A group always satisfies the schema without a placeholder member, and an empty
 group cannot exist. Removing a user or a member can need another step first. A later use of
 groups, such as allowing OpenVPN only to some, can read them as they are.
+
+## ADR-12: TOTP secrets are kept with the OATH-LDAP schema of slapd's otp module
+
+Status: Accepted
+
+Context: Users are enrolled for TOTP one-time passwords, and enrolled users must give a code
+whenever they log in, to OpenVPN and to `/auth`. Ubuntu's slapd ships the `otp` module, which
+holds the OATH-LDAP schema and an overlay that checks a code appended to the password at bind
+time; both OpenVPN (`openvpn-auth-ldap`) and `/auth` check passwords with a bind. Ubuntu
+packages `python3-pyotp`. An authenticator app is enrolled from a QR code of an `otpauth://` URI.
+
+Decision: The `otp` module is loaded and its overlay is configured on the package's database,
+after ppolicy, so every bind of an enrolled user takes the password followed by a current code
+and slapd records the time step used. The TOTP token is the user entry itself: enrolment adds `oathTOTPUser` and `oathTOTPToken` to it, with
+`oathTOTPToken` naming the entry, the secret in `oathSecret` and `oathTOTPParams` naming one
+shared parameter entry, `cn=totp,ou=policies`, with the parameters authenticator apps assume
+(6 digits, HMAC-SHA1, 30 seconds). An ACL lets the user manager account set `oathSecret` and
+no one read it. Enrolment is confirmed: `otp_new` makes a secret with `pyotp` and returns it
+with its URI, storing nothing, and `otp_enrol` stores the secret it is given back only when a
+code of it verifies, so a secret is never stored that the user's app does not produce. The
+router answers the URI; the client renders the QR code.
+
+Consequences: Removing a user removes its token. One check covers OpenVPN and `/auth`, and
+an enrolled user types the code after the password in the OpenVPN client. The secret passes
+through the client between `otp_new` and `otp_enrol`, so the router keeps no pending state;
+enrolling again replaces the secret only once a code of the new one verifies.
+
+## ADR-13: Members of admins manage the router; only CLI tokens change admins
+
+Status: Accepted
+
+Context: Signing in through `/auth` (ADR-3) is open to every user of the directory, most of
+whom only use OpenVPN. Finer permissions per operation are not needed. Somebody must be able to
+manage the router before any administrator exists and after every administrator lost access,
+and administrators must not be able to grant themselves or others the administrator role over
+the API.
+
+Decision: A token naming a user in `sub` may call the protected paths only while the user is a
+member of the group `admins`, read at each request; other users get 403. `vpn-router-token`, run
+as root on the router, makes tokens with the claim `admin: true` and a lifetime given at
+creation, without a maximum, signed with the router's key; `/auth` never issues the claim. An
+admin token may call everything, and only an admin token may change `admins`: create it, rename a
+group to it, change or remove it, add or take out its members, or remove a user who is a member
+of it. The first member is added with `vpn-router-group` on the router, which acts as root.
+
+Consequences: Removing a user from `admins` takes effect at the next request, without waiting
+for the user's tokens to expire. Membership of `admins` changes only through root on the router.
+An admin token is as powerful as root over the API and is revoked only by replacing the key, so
+it is given a short lifetime unless it is meant for automation.
+
+## ADR-14: Failed logins lock the account in slapd; /auth is rate limited per client
+
+Status: Accepted
+
+Context: `/auth` is open to the internal network and checks passwords, so it invites guessing;
+OpenVPN checks passwords from the internet. Both check with a bind, so slapd sees every attempt.
+ppolicy locks an account after a number of failed binds for a fixed time; it has no growing
+lockout. A Router Manager signs all its users in from one address.
+
+Decision: The default password policy locks an account for 15 minutes after 5 failed binds
+within 15 minutes; a successful bind clears the count; `vpn-router-user unlock` and
+`POST /users/<name>/unlock` end a lockout and clear the count, and leave a disabled user disabled.
+`/auth` answers the same 401 for a wrong name, password or code and for a locked or disabled
+user. `/auth` and `/auth/refresh` allow 10 calls a minute per client address, counted in memory,
+and answer 429 with `Retry-After` beyond that. The client address is the peer, or the last
+`X-Forwarded-For` entry for a peer in `web.trusted_proxies`, by default the address the service
+listens on, where a manager on the router calls from.
+
+Consequences: One lockout covers OpenVPN and `/auth`, kept by slapd alone. Someone who knows a
+user name can lock the user out for 15 minutes at a time. The rate limit restarts with the
+service. A manager elsewhere than on the router has its address added to `web.trusted_proxies`,
+or its users share one limit.
+
+## ADR-15: Refresh tokens are kept in PostgreSQL
+
+Status: Accepted
+
+Context: An access token lives 5 minutes, and a client stays signed in with a refresh token for
+8 hours; refresh tokens must be revocable when a user signs out, is disabled, changes the
+password or is removed, and must survive a restart of the service. The Router Manager keeps its
+sessions in PostgreSQL too.
+
+Decision: Refresh tokens are random strings, stored as their SHA-256 with the user name and the
+expiry in the table `refresh_tokens`, through psycopg 3 (`python3-psycopg`). The database is
+`web.database`, or by default the database and role `vpn-router` on the router, which the web
+service creates on its first run (ADR-16) with `runuser -u postgres -- psql` and a random password in
+`/etc/vpn-router/database.secret`, reached over TCP on `localhost`, as Ubuntu's `pg_hba.conf`
+takes peer authentication on the local socket. A refresh token is used once: refreshing deletes
+it and issues a new one, after checking that the user is neither disabled nor locked.
+
+Consequences: The package depends on `postgresql` and `python3-psycopg`. A stolen database holds
+no usable token. Changing a user's password, disabling, unlocking or removing the user, and
+signing out, end the user's sessions at once; an access token already issued lives out its
+5 minutes.
+
+## ADR-16: Each service creates what it needs on its first run
+
+Status: Accepted
+
+Context: The package manager configures packages in an order of its own; on a new VM the Router
+Manager was configured before the PostgreSQL server, so a database created at installation was
+missing. PostgreSQL may also start after the services at boot.
+
+Decision: The web service makes its signing key and creates the refresh tokens' database itself,
+and the Router Manager its sessions database; neither `vpn-router-setup` nor a maintainer script
+creates them. Each service does it before it listens, retrying every 15 seconds and logging each
+failure, then listens. The first step is the PostgreSQL driver: psycopg finds libpq through the
+linker cache, which the package manager refreshes only at the end of an installation, so a
+service started during the installation cannot import it yet. The Manager imports psycopg only in
+its retry loop. On the router, `vpnrouter_auth` alone imports it, in `driver()`, which the web
+service's retry loop calls first, and turns a failed import and every database failure into
+`DatabaseError`, the only database exception the web service and `vpn-router-user` handle.
+
+Consequences: The order of installation and of the start of PostgreSQL and of the linker cache
+does not matter; a driver that cannot be loaded or a database that cannot be created is retried
+until the cause is gone, and the reason is in the service's log, without restarts. The web unit may write `/etc/vpn-router` for the key and the database password.
+
+## ADR-17: Bootstrap diagnostics run only until the first setup succeeds
+
+Status: Accepted
+
+Context: A new VM is reachable only once its setup has worked, and the reason of a failure must
+be readable from the internal network before an SSH session is possible. A service that reports
+the setup cannot be one of the services being set up: those import packages that may not load yet
+(psycopg could not find libpq before `ldconfig` ran), and they would fail before reporting
+anything. Once the router has been set up, SSH is the way to diagnose it. The more a diagnostic
+page does, the more ways it has to fail itself.
+
+Decision: Each package ships a bootstrap service of the Python standard library alone, a script
+that answers every request with the plain-text output of a fixed list of commands: the setup
+state, the services, the failed units, the addresses and the end of the logs. The router's
+listens on 17081 and the Router Manager's on 17080, beside the ports of the services, 7081 and
+7080, on every address. The units are never enabled and not started at installation;
+`vpn-router-setup` starts them while `/var/lib/vpn-router/setup.done` does not exist, and opens
+their ports on the internal interface. The first successful run writes that file, closes the ports
+and stops them; a failed run leaves them running, so the reason stays readable. Setup records its
+state, step and failure in `/run/vpn-router/setup.json` at every step; `vpnrouter.fail` ends with
+a `SystemExit` carrying its message, so setup records it.
+
+Consequences: The diagnostics need nothing the setup provides and cannot collide with the real
+services. After the first success they never run again, also not after an upgrade or a reboot.
+The page shows log lines to whoever reaches the internal network during the first setup.
@@ -86,8 +86,8 @@ that deploy a router on Azure or GCP.

 `/etc/vpn-router/vpn-router.json` is JSON, mode 0600, and is created on first install. It
 is managed with `ucf`, so it is yours to edit: a later `dpkg-reconfigure` applies the answers
-given and asks before touching anything you altered by hand, and an upgrade leaves the file as
-it is and asks nothing. Settings the file leaves out take their defaults when it is read, so a
+given and asks before touching anything you altered by hand, and an upgrade asks no questions
+and leaves the file as it is. Settings the file leaves out take their defaults when it is read, so a
 file written by cloud-init or by hand needs only the settings that differ. A complete example is
 installed at `/usr/share/doc/vpn-router/vpn-router.json.example`.

@@ -125,7 +125,8 @@ must have the type shown below; an unknown section or setting is an error.
 | `docker` | `exposed_ports` | list | | TCP ports and port ranges (`8000:8100`) reachable from the external interface; at most 15, a range counting as two |
 | `web` | `enabled` | boolean | `true` | The status web service on the internal address |
 | `web` | `port` | integer | `7081` | TCP port of the status web service |
-| `web` | `jwt_secret` | string | | Secret, at least 32 characters, that signs the tokens for `/config`; without it `/config` is not served |
+| `web` | `database` | string | | A PostgreSQL connection string for the refresh tokens; without it, the `vpn-router` database on this host that the web service creates |
+| `web` | `trusted_proxies` | list | the listening address | Addresses whose `X-Forwarded-For` header names the client for the `/auth` rate limit |

 A minimal site-to-site configuration in `auto` mode:

@@ -310,7 +311,7 @@ certificate names:
 - the certificate is checked against the CA and the CRL (`crl-verify`, read at every
  connection);
 - the password is checked by `openvpn-auth-ldap`, which binds to the user directory as the
-  user;
+  user; a user enrolled for TOTP types the password followed by the current 6-digit code;
 - `/usr/lib/vpn-router/openvpn-same-user` refuses a login whose user name is not the common
  name of the certificate.

@@ -331,6 +332,14 @@ is switched on:
  `/etc/vpn-router/ldap/user-manager.secret` (root only);
 - passwords are stored as ARGON2 hashes, and a disabled user is locked with the password policy
  (`pwdAccountLockedTime`);
+- the password policy locks an account for 15 minutes after 5 failed binds within 15 minutes,
+  whether they come from OpenVPN or from `/auth`; a successful bind clears the count, and
+  `vpn-router-user unlock` ends a lockout;
+- slapd's `otp` overlay is configured on the database; a user enrolled for TOTP is also an
+  `oathTOTPUser` and `oathTOTPToken` whose `oathTOTPToken` names the entry itself, with the
+  secret in `oathSecret`, which the user manager account may set and no one can read, and
+  `oathTOTPParams` naming `cn=totp,ou=policies` (6 digits, HMAC-SHA1, 30 second steps, one step
+  of tolerance). A bind of an enrolled user takes the password followed by a current code;
 - slapd listens on its local socket only (`SLAPD_SERVICES` in `/etc/default/slapd`).

 The directory is kept on purge.
@@ -340,8 +349,9 @@ The directory is kept on purge.
 `vpn-router-user` manages the users. A user is the directory entry, with a first name, a last
 name and an e-mail address (`givenName`, `sn` and `mail`; `cn` is the first and the last name),
 and a client certificate with the user name as its common name, in `/etc/vpn-router/pki/users`.
-Passwords are read from the first line of standard input. The answer is text: a table of the
-users, the client profile or a short message, and a failure is a message on standard error.
+Passwords and TOTP secrets are read from the first line of standard input. The answer is text: a
+table of the users, the client profile, a TOTP secret or a short message, and a failure is a
+message on standard error.
 `--json`, before the command, gives the answer as JSON on standard output, and a failure as
 `{"error": ...}` there too; the web service uses it. `-h` or `--help` shows the commands, and
 after a command its options.
@@ -355,22 +365,32 @@ vpn-router-user --json list
 echo 'n3w-pass' | vpn-router-user passwd alice
 vpn-router-user disable alice
 vpn-router-user enable alice
+vpn-router-user unlock alice
 vpn-router-user remove alice
+vpn-router-user otp-new alice
+echo 'JBSWY3DPEHPK3PXPJBSWY3DPEHPK3PXP' | vpn-router-user otp-enrol alice 123456
+vpn-router-user otp-remove alice
 ```

 | Command | Effect |
 | --- | --- |
-| `list` | The users with their names and e-mail addresses, whether each is enabled and has a certificate |
+| `list` | The users with their names and e-mail addresses, whether each is enabled or locked, has a certificate and is enrolled for TOTP |
 | `add <name> --first-name F --last-name L --email E` | Adds the user and issues the certificate; all three are required |
 | `update <name> [--first-name F] [--last-name L] [--email E]` | Changes the names or the e-mail address |
 | `passwd <name>` | Sets the password |
-| `disable <name>`, `enable <name>` | Locks the user out of OpenVPN, and lets the user in again |
+| `disable <name>`, `enable <name>` | Locks the user out of OpenVPN and `/auth`, and lets the user in again |
+| `unlock <name>` | Ends a lockout after failed logins and clears the failure count; a disabled user stays disabled |
 | `remove <name>` | Takes the user out of its groups, revokes the certificate, renews the CRL and removes the user; refused while the user is the last member of a group |
 | `profile <name>` | The OpenVPN client profile, with the CA, the certificate and the key inline |
+| `otp-new <name>` | A new TOTP secret and its `otpauth://` URI, issued by `wan.local_fqdn`, for an authenticator app; nothing is stored |
+| `otp-enrol <name> <code>` | Stores the secret read from standard input when the code is a current code of it, replacing a secret the user has |
+| `otp-remove <name>` | Removes the user's TOTP secret |

 A user name is 1 to 32 lowercase letters, digits and hyphens, starting with a letter; a
 password has 8 to 128 characters; a first or last name has 1 to 64 printable characters; an
-e-mail address has one `@`, no spaces and a dot in its domain. In the answers a user added
+e-mail address has one `@`, no spaces and a dot in its domain; a TOTP secret is 32 base32
+characters, as `otp-new` makes it, and a code 6 digits. Once enrolled, a user logs in to OpenVPN
+and `/auth` with the password followed by a current code. In the answers a user added
 without the names or the e-mail address has them as `null`. The exit status is 0 on success, 2 for an invalid request,
 3 when the user does not exist, 4 when the user exists and 1 for any other failure.

@@ -453,10 +473,32 @@ The `vpn-router-web` service shows the state of the router on its internal addre
 `web.port` (7081 by default). Its pages `/` and `/status` need no authentication and show no
 configuration: no key, identity, address or subnet.

+The service first makes its signing key and the refresh tokens' database, retrying every 15
+seconds, for example while PostgreSQL is not running yet, and logs each failure; then it listens.
+
+### Bootstrap diagnostics
+
+Until a run of `vpn-router-setup` has succeeded, every run starts `vpn-router-bootstrap`, which
+serves diagnostics as plain text on port 17081 of every address, and the firewall opens 17081 and
+17080 (the Router Manager's bootstrap) on the internal interface. The first successful run stops
+the bootstrap services and closes their ports, and writes `/var/lib/vpn-router/setup.done`, after
+which they never run again; a failed run leaves them running, so the reason stays readable until
+a later run succeeds. The unit is never enabled.
+
+The page, read again every 10 seconds, shows the state, the step and the failure of the last
+setup run (`/run/vpn-router/setup.json`, which setup writes at every step), the state of the
+router's services, the failed units, the addresses and routes, `cloud-init status`, and the end of
+the logs of `vpn-router-setup` and `vpn-router-web`:
+
+```bash
+curl http://10.0.0.4:17081/
+```
+
 | Page | Content |
 | --- | --- |
 | `/` | The status for people: plain text for `curl` and `wget`, a page styled with Tailwind CSS for a browser, in the browser's light or dark scheme, that reloads itself every minute (the style sheet is loaded from a CDN, so the browser needs internet access; the text is readable without it) |
 | `/status` | The same status as JSON |
+| `/auth` | Signing in with a user's password and TOTP code, refreshing and signing out |
 | `/config` | The configuration without its secret options, and what the router resolved and runs, as JSON; needs a token |
 | `/users` | The users of OpenVPN and road-warrior access; needs a token |
 | `/groups` | The groups of users; needs a token |
@@ -476,43 +518,70 @@ curl http://10.0.0.4:7081/
 curl http://10.0.0.4:7081/status
 ```

+### Tokens and signing in
+
+`/config`, `/users` and `/groups` need a JWT in an `Authorization: Bearer` header, signed with
+HS256 by the key the web service makes on its first run, `/etc/vpn-router/web.key` (root only).
+The key never leaves the router; while it does not exist these paths and `/auth` are not served
+(404). A missing, expired or wrongly signed token, or one of another algorithm, gets 401 with
+`WWW-Authenticate: Bearer`. There are two kinds of tokens:
+
+- an access token of `/auth` names a user in `sub` and is valid for 5 minutes; the user must
+  be a member of the group `admins`, otherwise the request gets 403;
+- a token of `vpn-router-token`, run as root on the router, has the claim `admin` and is valid
+  for the lifetime given, without a maximum. It may call everything, and only such a token may
+  change the group `admins`: create it, rename a group to it, change or remove it, add or take
+  out its members, or remove a user who is a member of it.
+
+The first member of `admins` is added on the router:
+
+```sh
+vpn-router-group add admins --member alice --description 'Router administrators'
+vpn-router-token --hours 1
+vpn-router-token --json --days 30
+```
+
+`POST /auth` signs a user in. Request and answer bodies are JSON:
+
+| Request | Body | Answer |
+| --- | --- | --- |
+| `POST /auth` | `{"name", "password", "code"}`; `code` only for a user enrolled for TOTP | the tokens |
+| `POST /auth/refresh` | `{"refresh_token"}` | new tokens; the refresh token given is used up |
+| `POST /auth/logout` | `{"refresh_token"}` | `{"revoked": true}`, or `false` for a token not known |
+
+The tokens are `{"access_token", "token_type": "Bearer", "expires_in": 300, "refresh_token",
+"refresh_expires_in": 28800}`. A wrong name, password or code, and a locked or disabled user,
+get 401 with the same message; a refresh token is refused once its user is disabled, locked or
+removed. Changing a user's password, disabling, unlocking or removing the user revokes the
+user's refresh tokens. Five failed logins lock the account for 15 minutes (see the user
+directory). `/auth` and `/auth/refresh` take at most 10 calls a minute from one client address,
+and answer 429 with `Retry-After` beyond that; for a call from one of `web.trusted_proxies`, by
+default the address the service listens on, the client address is the last entry of
+`X-Forwarded-For`.
+
+Refresh tokens are kept, as their SHA-256, in PostgreSQL: by default in the database and role
+`vpn-router` on the router, which the web service creates on its first run with a password in
+`/etc/vpn-router/database.secret` (root only), reached over TCP on `localhost`; `web.database`
+names another database. Tokens and passwords travel over plain HTTP on the internal network.
+
+```bash
+curl -H 'Content-Type: application/json' -d '{"name": "alice", "password": "s3cr3t-pass", "code": "123456"}' \
+     http://10.0.0.4:7081/auth
+curl -H "Authorization: Bearer $TOKEN" http://10.0.0.4:7081/config
+```
+
 ### The /config page

 `/config` shows what `/` and `/status` leave out, for a client such as manage-ui:

 | Key | Content |
 | --- | --- |
-| `configuration` | `vpn-router.json` with the defaults applied, without its secret options (`psk`, `jwt_secret`, and any option named like a password, token or key) |
+| `configuration` | `vpn-router.json` with the defaults applied, without its secret options (`psk`, `database`, and any option named like a secret, password, token or key) |
 | `network` | The external and internal interface, their addresses, the internal gateway, the local and remote IKE identities and subnets, and the IPv4 routes as `ip -j -4 route` prints them |
 | `ipsec` | `connections` and `sas`: one object per connection and per IKE SA, read from `swanctl --list-conns --pretty` and `swanctl --list-sas --pretty`, with the connection's name under `name` and swanctl's keys and string values |
 | `wireguard` | The public key and port of `wg0`, and each peer's public key, endpoint, allowed IPs, handshake age and traffic; no private or preshared key |
 | `docker` | `docker.exposed_ports` and the rules of the `VPN-ROUTER-DOCKER` chain |

-The page needs a JWT in an `Authorization: Bearer` header, signed with HS256 by `web.jwt_secret`.
-The token must have an `exp` claim, not more than ten minutes from now (a minute of clock
-difference is allowed); `nbf` is checked when present. Other algorithms are refused, with 401 and
-`WWW-Authenticate: Bearer`. While `web.jwt_secret` is not set the page is not served (404). The
-token and the page travel over plain HTTP on the internal network, and the token can be used
-until it expires, so keep the lifetime short. A client makes a token from the secret with only
-HMAC-SHA256; with Python:
-
-```python
-import base64, hashlib, hmac, json, time
-
-def b64(data):
-    return base64.urlsafe_b64encode(data).rstrip(b'=').decode()
-
-secret = 'the web.jwt_secret value'
-head = b64(json.dumps({'alg': 'HS256', 'typ': 'JWT'}).encode())
-body = b64(json.dumps({'exp': int(time.time()) + 300}).encode())
-sign = b64(hmac.new(secret.encode(), f'{head}.{body}'.encode(), hashlib.sha256).digest())
-print(f'{head}.{body}.{sign}')
-```
-
-```bash
-curl -H "Authorization: Bearer $TOKEN" http://10.0.0.4:7081/config
-```
-
 ### The /users API

 `/users` manages users with the same token as `/config`, with the module `vpn-router-user` uses
@@ -521,19 +590,23 @@ and answer bodies are JSON; a request body is at most 4096 bytes.

 | Request | Body | Answer |
 | --- | --- | --- |
-| `GET /users` | | `{"users": [{"name", "enabled", "certificate", "first_name", "last_name", "email"}, ...]}` |
+| `GET /users` | | `{"users": [{"name", "enabled", "locked", "certificate", "otp", "first_name", "last_name", "email"}, ...]}` |
 | `POST /users` | `{"name", "password", "first_name", "last_name", "email"}`, all required | 201 and the user |
 | `PUT /users/<name>` | any of `{"first_name", "last_name", "email"}` | the user |
 | `PUT /users/<name>/password` | `{"password": ...}` | the user |
 | `POST /users/<name>/disable` | | the user |
 | `POST /users/<name>/enable` | | the user |
+| `POST /users/<name>/unlock` | | the user |
 | `DELETE /users/<name>` | | `{"name", "removed": true}` |
 | `GET /users/<name>/profile` | | `{"name", "profile"}`, the OpenVPN client profile |
+| `POST /users/<name>/otp` | | `{"name", "secret", "uri"}`, a new TOTP secret; nothing is stored |
+| `PUT /users/<name>/otp` | `{"secret", "code"}` | the user, once the code matches the secret |
+| `DELETE /users/<name>/otp` | | the user |

 A failure is `{"error": ...}` with 400 for an invalid request, 404 for a user that does not
 exist, 409 for a user that exists and 500 otherwise. `/users` works whether or not OpenVPN or
-road-warrior access is enabled; the profile needs OpenVPN. Passwords and private keys cross the internal
-network in plain HTTP, like the token.
+road-warrior access is enabled; the profile needs OpenVPN. Passwords, private keys and TOTP
+secrets cross the internal network in plain HTTP, like the token.

 ```bash
 curl -H "Authorization: Bearer $TOKEN" -H 'Content-Type: application/json' \
@@ -563,7 +636,7 @@ The service starts on its own, without waiting for `vpn-router-setup`, and keeps
 30 seconds it works out the internal address from `vpn-router.json` as `general.mode` directs,
 and reads `web.port` and `web.enabled`; it listens there, moves when the address or the port
 changes, and stops listening while `web.enabled` is `false` or the address is not known. Each
-request reads the token secret from `vpn-router.json`, and the features and the facts for
+request reads the signing key and `vpn-router.json`, and the features and the facts for
 `/config` from `/run/vpn-router/web.json` (readable by root only), which `vpn-router-setup` writes;
 before that file exists the page shows no features. `vpn-router-setup` opens `web.port` on the
 internal interface in the firewall.
@@ -586,8 +659,9 @@ may be in use by something else.

 `apt-get purge` additionally removes the generated files and the firewall rules, restoring
 `/etc/ufw/before.rules` to its original content, and hands `vpn-router.json` back to `ucf`
-before deleting it so a later reinstall starts clean. Key material and users are never
-deleted: `/etc/vpn-router/pki`, `/etc/vpn-router/ldap`, the user directory in slapd, the
+before deleting it so a later reinstall starts clean. It also removes the web service's
+signing key, and the refresh tokens' database, its role and their password file when the package
+created them. Key material and users are never deleted: `/etc/vpn-router/pki`, `/etc/vpn-router/ldap`, the user directory in slapd, the
 WireGuard key pair and everything under `/etc/swanctl` are left alone.

 ## Building
@@ -1,3 +1,109 @@
+vpn-router (0.9.1-1) unstable; urgency=medium
+
+  fix: the web service waits for the PostgreSQL driver instead of failing at start.
+
+  * The web service imports psycopg as the first step of its
+    initialization, retrying every 15 seconds and logging the failure while
+    libpq cannot be found yet during an installation.
+  * vpn-router-user and vpn-router-token load without psycopg; a command
+    that needs the token database then reports that the driver cannot be
+    loaded, as it reports every failure of that database.
+
+ -- Slawomir Koszewski <slawek@koszewscy.waw.pl>  Sun, 25 Oct 2026 05:00:00 +0200
+
+vpn-router (0.9.0-1) unstable; urgency=medium
+
+  feature: bootstrap diagnostics on port 17081 during the first setup.
+
+  * vpn-router-bootstrap serves, on port 17081, the setup state, the
+    services, the failed units, the addresses, cloud-init status and the end
+    of the logs of setup and the web service, as plain text.
+  * vpn-router-setup starts it, and the Router Manager's bootstrap on port
+    17080, until a run has succeeded, and opens both ports on the internal
+    interface; the first successful run writes
+    /var/lib/vpn-router/setup.done, closes the ports and stops them. A
+    failed run leaves them running.
+  * The web service no longer has a starting phase: it creates its key and
+    database, then listens.
+
+ -- Slawomir Koszewski <slawek@koszewscy.waw.pl>  Sun, 25 Oct 2026 04:00:00 +0200
+
+vpn-router (0.8.0-1) unstable; urgency=medium
+
+  feature: the web service creates its key and database itself and shows its starting phase.
+
+  * vpn-router-web makes its signing key and creates the refresh tokens'
+    database on its first run, retrying every 15 seconds until PostgreSQL
+    answers; vpn-router-setup no longer creates them.
+  * Until then / and /status show the web service's phase and its failure,
+    with the state, the step and the failure of the last run of
+    vpn-router-setup, and every page answers 503.
+  * vpn-router-setup records its state, current step and failure in
+    /run/vpn-router/setup.json.
+
+ -- Slawomir Koszewski <slawek@koszewscy.waw.pl>  Sun, 25 Oct 2026 03:00:00 +0200
+
+vpn-router (0.7.1-1) unstable; urgency=medium
+
+  fix: a failing user directory or token database no longer leaves the tunnels down.
+
+  * vpn-router-setup brings up strongSwan and WireGuard before it provisions
+    the user directory and the token database. A failure of either no
+    longer stops the other steps; setup reports it and fails at the end.
+
+ -- Slawomir Koszewski <slawek@koszewscy.waw.pl>  Sun, 25 Oct 2026 02:00:00 +0200
+
+vpn-router (0.7.0-1) unstable; urgency=medium
+
+  feature: users sign in to the web service, and only members of admins manage the router.
+
+  * POST /auth signs a user in with the password, and the TOTP code for an
+    enrolled user, and answers a 5 minute access token and an 8 hour refresh
+    token; /auth/refresh renews them and /auth/logout revokes the refresh
+    token.
+  * /config, /users and /groups take tokens of users in the group admins,
+    and tokens of the new vpn-router-token, which has an admin claim and a
+    chosen lifetime. Only these may change the group admins.
+  * The tokens are signed with /etc/vpn-router/web.key, which
+    vpn-router-setup makes on the router; web.jwt_secret is ignored.
+  * Five failed logins lock an account for 15 minutes, for OpenVPN and
+    /auth. vpn-router-user unlock and POST /users/<name>/unlock end the
+    lockout, and the users show locked.
+  * /auth takes at most 10 calls a minute per client address; for calls from
+    web.trusted_proxies the address is read from X-Forwarded-For.
+  * Users enrolled for TOTP log in to OpenVPN with the password followed by
+    the code.
+  * Refresh tokens are kept in PostgreSQL, in the database web.database
+    names or in the vpn-router database vpn-router-setup creates. The
+    package depends on postgresql and python3-psycopg.
+
+ -- Slawomir Koszewski <slawek@koszewscy.waw.pl>  Sun, 25 Oct 2026 01:00:00 +0200
+
+vpn-router (0.6.0-1) unstable; urgency=medium
+
+  feature: users can be enrolled for TOTP one-time passwords.
+
+  * vpn-router-user otp-new makes a TOTP secret and its otpauth URI,
+    otp-enrol stores the secret, read from standard input, given a code of
+    it, and otp-remove removes it. list and every answer show whether a user
+    is enrolled, as otp.
+  * The web service enrols with POST, PUT and DELETE /users/<name>/otp.
+  * The secret is kept in the user's entry with the OATH-LDAP schema of
+    slapd's otp module; only the user manager account may set it, and no
+    one can read it. Logins do not ask for the codes.
+  * The package depends on python3-pyotp.
+
+ -- Slawomir Koszewski <slawek@koszewscy.waw.pl>  Sat, 24 Oct 2026 23:30:00 +0200
+
+vpn-router (0.5.9-1) unstable; urgency=medium
+
+  fix: package upgrades no longer ask the configuration questions.
+
+  * The configuration questions are asked only on a fresh installation and
+    by dpkg-reconfigure.
+
+ -- Slawomir Koszewski <slawek@koszewscy.waw.pl>  Sat, 24 Oct 2026 21:00:00 +0200
+
 vpn-router (0.5.8-1) unstable; urgency=medium

  fix: an upgrade keeps the configuration and asks nothing.
@@ -2,6 +2,12 @@
 set -e
 . /usr/share/debconf/confmodule

+# An upgrade (configure with the old version) asks nothing; vpn-router.json holds the
+# configuration, and dpkg-reconfigure asks the questions again.
+if [ "$1" = "configure" ] && [ -n "$2" ]; then
+    exit 0
+fi
+
 # Every question has a default, so a non-interactive install with nothing
 # preseeded completes without prompting. The defaults are platform "none" and
 # mode "manual", which together mean: install the files, configure nothing.
@@ -21,6 +21,9 @@ Depends: ${misc:Depends},
         python3,
         python3-jinja2,
         python3-ldap,
+         python3-pyotp,
+         python3-psycopg,
+         postgresql,
         slapd,
         openvpn,
         openvpn-auth-ldap,
@@ -2,15 +2,19 @@ src/etc/sysctl.d/99-vpn-router.conf                   etc/sysctl.d/
 src/usr/lib/vpn-router/vpnrouter.py                   usr/lib/vpn-router/
 src/usr/lib/vpn-router/vpnrouter_ldap.py              usr/lib/vpn-router/
 src/usr/lib/vpn-router/vpnrouter_net.py               usr/lib/vpn-router/
+src/usr/lib/vpn-router/vpnrouter_serve.py             usr/lib/vpn-router/
 src/usr/lib/vpn-router/vpnrouter_users.py             usr/lib/vpn-router/
+src/usr/lib/vpn-router/vpnrouter_auth.py              usr/lib/vpn-router/
 src/usr/lib/vpn-router/openvpn-same-user              usr/lib/vpn-router/
 src/usr/sbin/vpn-router-user                          usr/sbin/
 src/usr/sbin/vpn-router-group                         usr/sbin/
+src/usr/sbin/vpn-router-token                         usr/sbin/
 src/usr/lib/vpn-router/simple-ca                      usr/lib/vpn-router/
 src/usr/lib/vpn-router/configure                      usr/lib/vpn-router/
 src/usr/lib/vpn-router/docker-rules                   usr/lib/vpn-router/
 src/usr/lib/vpn-router/setup                          usr/lib/vpn-router/
 src/usr/lib/vpn-router/web                            usr/lib/vpn-router/
+src/usr/lib/vpn-router/bootstrap                      usr/lib/vpn-router/
 src/usr/lib/vpn-router/generate-config                usr/lib/vpn-router/
 src/usr/lib/vpn-router/vpnrouter_platforms/*.py       usr/lib/vpn-router/vpnrouter_platforms/
 src/usr/share/vpn-router/templates/*                  usr/share/vpn-router/templates/
@@ -11,10 +11,10 @@ strip_ufw_blocks() {

    tmp="$(mktemp)"
    awk '
-        /^# (IPSEC|WIREGUARD|OPENVPN|P2S DNS|WEB|MANAGER|ROUTER FORWARD|ROUTER NAT|ROUTER MSS) RULES START$/ {
+        /^# (IPSEC|WIREGUARD|OPENVPN|P2S DNS|WEB|BOOTSTRAP|MANAGER|ROUTER FORWARD|ROUTER NAT|ROUTER MSS) RULES START$/ {
            skip = 1; pending = 0; next
        }
-        /^# (IPSEC|WIREGUARD|OPENVPN|P2S DNS|WEB|MANAGER|ROUTER FORWARD|ROUTER NAT|ROUTER MSS) RULES END$/ {
+        /^# (IPSEC|WIREGUARD|OPENVPN|P2S DNS|WEB|BOOTSTRAP|MANAGER|ROUTER FORWARD|ROUTER NAT|ROUTER MSS) RULES END$/ {
            skip = 0; next
        }
        skip { next }
@@ -49,7 +49,17 @@ case "$1" in
              /etc/wireguard/wg0.conf \
              /etc/openvpn/server/vpn-router.conf \
              /etc/openvpn/server/vpn-router-ldap.conf \
-              /etc/vpn-router/vpn-router.json
+              /etc/vpn-router/vpn-router.json \
+              /etc/vpn-router/web.key \
+              /var/lib/vpn-router/setup.done
+        rmdir --ignore-fail-on-non-empty /var/lib/vpn-router 2>/dev/null || true
+
+        # The refresh tokens' database and role, when the package created them.
+        if [ -f /etc/vpn-router/database.secret ] && command -v psql >/dev/null 2>&1; then
+            runuser -u postgres -- psql -X -q -c 'DROP DATABASE IF EXISTS "vpn-router"' || true
+            runuser -u postgres -- psql -X -q -c 'DROP ROLE IF EXISTS "vpn-router"' || true
+        fi
+        rm -f /etc/vpn-router/database.secret

        # Key material and users are not package state: /etc/vpn-router/pki,
        # /etc/vpn-router/ldap, the user directory in slapd, the WireGuard key
@@ -6,3 +6,4 @@ override_dh_installsystemd:
 	dh_installsystemd --name=vpn-router-setup
 	dh_installsystemd --name=vpn-router-docker
 	dh_installsystemd --name=vpn-router-web
+	dh_installsystemd --name=vpn-router-bootstrap --no-enable --no-start
@@ -0,0 +1,14 @@
+[Unit]
+Description=VPN Router bootstrap diagnostics page
+Documentation=file:/usr/share/doc/vpn-router/README.md
+
+# Never enabled: vpn-router-setup starts it until a setup has succeeded, and stops it after that.
+[Service]
+Type=simple
+ExecStart=/usr/bin/python3 /usr/lib/vpn-router/bootstrap
+Restart=on-failure
+RestartSec=5
+NoNewPrivileges=yes
+ProtectSystem=strict
+ProtectHome=yes
+PrivateTmp=yes
@@ -1,8 +1,8 @@
 [Unit]
 Description=VPN Router Setup
 Documentation=file:/usr/share/doc/vpn-router/README.md
-Wants=network-online.target
-After=network-online.target docker.service
+Wants=network-online.target postgresql.service
+After=network-online.target docker.service postgresql.service

 [Service]
 Type=oneshot
@@ -1,7 +1,7 @@
 [Unit]
 Description=VPN Router status web service
 Documentation=file:/usr/share/doc/vpn-router/README.md
-After=network-online.target
+After=network-online.target postgresql.service
 Wants=network-online.target

 [Service]
@@ -11,7 +11,7 @@ Restart=on-failure
 RestartSec=15
 NoNewPrivileges=yes
 ProtectSystem=strict
-ReadWritePaths=-/run/vpn-router -/etc/vpn-router/pki -/etc/swanctl/x509crl
+ReadWritePaths=-/run/vpn-router -/etc/vpn-router -/etc/swanctl/x509crl
 ProtectHome=yes
 PrivateTmp=yes

@@ -0,0 +1,55 @@
+#!/usr/bin/python3
+"""Serve diagnostics of the router's first setup on PORT, with the Python standard library alone.
+
+Run by vpn-router-bootstrap.service, which vpn-router-setup starts until a setup has succeeded and
+then stops. Every request on any address answers, as plain text, the output of COMMANDS run at that
+moment: the state of the last setup run, the state of the services, the addresses, and the end of
+the logs.
+"""
+
+import subprocess
+from http.server import BaseHTTPRequestHandler, ThreadingHTTPServer
+
+PORT = 17081
+
+#: The title of each section and its command.
+COMMANDS = (
+    ('Setup state', ['cat', '/run/vpn-router/setup.json']),
+    ('Services', ['systemctl', 'is-active', 'vpn-router-setup', 'vpn-router-web', 'slapd', 'postgresql',
+                  'strongswan', 'openvpn-server@vpn-router', 'wg-quick@wg0']),
+    ('Failed units', ['systemctl', '--failed', '--no-legend', '--plain']),
+    ('Addresses', ['ip', '-brief', 'address']),
+    ('Routes', ['ip', 'route']),
+    ('cloud-init', ['cloud-init', 'status', '--long']),
+    ('vpn-router-setup log', ['journalctl', '--unit', 'vpn-router-setup', '--lines', '60', '--no-pager', '--output', 'short-iso']),
+    ('vpn-router-web log', ['journalctl', '--unit', 'vpn-router-web', '--lines', '30', '--no-pager', '--output', 'short-iso']),
+)
+
+
+def report():
+    """The output of every command of COMMANDS under its title."""
+    sections = []
+    for title, command in COMMANDS:
+        try:
+            result = subprocess.run(command, capture_output=True, text=True, timeout=10)
+            text = (result.stdout + result.stderr).rstrip()
+        except (OSError, subprocess.TimeoutExpired) as exc:
+            text = f'cannot run {command[0]}: {exc}'
+        sections.append(f'== {title} ==\n{text}\n')
+    return '\n'.join(sections)
+
+
+class Handler(BaseHTTPRequestHandler):
+    def do_GET(self):
+        body = report().encode()
+        self.send_response(200)
+        self.send_header('Content-Type', 'text/plain; charset=utf-8')
+        self.send_header('Content-Length', str(len(body)))
+        self.send_header('Cache-Control', 'no-store')
+        self.send_header('Refresh', '10')
+        self.end_headers()
+        self.wfile.write(body)
+
+
+if __name__ == '__main__':
+    ThreadingHTTPServer(('', PORT), Handler).serve_forever()
@@ -5,6 +5,12 @@ Regenerates the configuration files first, then does everything that needs
 the live system: routes, DNS-derived identity, WireGuard keys, firewall rules
 and service reloads. Idempotent - a second run changes nothing.

+Until a run has succeeded (SETUP_DONE), every run starts the bootstrap
+services, BOOTSTRAP_UNITS, which show the progress and the failure recorded in
+SETUP_STATE on BOOTSTRAP_PORTS, and the firewall opens those ports on the
+internal interface. The first successful run writes SETUP_DONE, closes the
+ports and stops the bootstrap services; a failed run leaves them running.
+
 Interface names come from the configuration and are never guessed. The
 firewall and NAT rules are written against them, so choosing the wrong one
 would be a silent and security-relevant mistake.
@@ -17,11 +23,10 @@ import re
 import socket
 import subprocess
 import sys
+import time

 sys.path.insert(0, os.path.dirname(os.path.realpath(__file__)))

-import ldap
-
 import vpnrouter
 import vpnrouter_ldap
 from vpnrouter import fail
@@ -31,6 +36,16 @@ UFW_DEFAULTS = pathlib.Path('/etc/default/ufw')
 WG_DIR       = pathlib.Path('/etc/wireguard')
 WEB_STATE    = pathlib.Path('/run/vpn-router/web.json')

+#: The state, the current step and the failure of the last run, which the bootstrap services show.
+SETUP_STATE  = pathlib.Path('/run/vpn-router/setup.json')
+
+#: Written once a run has succeeded; until then setup runs the bootstrap services and opens their ports.
+SETUP_DONE   = pathlib.Path('/var/lib/vpn-router/setup.done')
+
+#: The units of the bootstrap services, and their ports: the Router Manager's and the router's.
+BOOTSTRAP_UNITS = ('vpn-router-bootstrap.service', 'vpn-router-manager-bootstrap.service')
+BOOTSTRAP_PORTS = (17080, 17081)
+
 #: The Router Manager, vpn-router-manager, when it is installed, and its settings.
 MANAGER        = pathlib.Path('/usr/lib/vpn-router-manager/manager')
 MANAGER_CONFIG = pathlib.Path('/etc/vpn-router-manager/manager.json')
@@ -39,7 +54,7 @@ MANAGER_PORT   = 7080
 FILTER_ANCHOR = '# End required lines'

 #: Configuration options left out wherever the configuration is shown.
-SECRET_OPTION = re.compile(r'(^psk$)|secret|password|token|_key$')
+SECRET_OPTION = re.compile(r'(^psk$)|secret|password|token|_key$|^database$')

 BLOCKS = (
    ('# IPSEC RULES START',         '# IPSEC RULES END'),
@@ -47,6 +62,7 @@ BLOCKS = (
    ('# OPENVPN RULES START',       '# OPENVPN RULES END'),
    ('# P2S DNS RULES START',       '# P2S DNS RULES END'),
    ('# WEB RULES START',           '# WEB RULES END'),
+    ('# BOOTSTRAP RULES START',     '# BOOTSTRAP RULES END'),
    ('# MANAGER RULES START',       '# MANAGER RULES END'),
    ('# ROUTER FORWARD RULES START', '# ROUTER FORWARD RULES END'),
    ('# ROUTER NAT RULES START',    '# ROUTER NAT RULES END'),
@@ -200,6 +216,14 @@ def filter_block(ctx):
            '# WEB RULES END',
        ]

+    if not SETUP_DONE.exists() and ctx['int_iface']:
+        lines += [
+            '',
+            '# BOOTSTRAP RULES START',
+            *(f'-A ufw-before-input -i {ctx["int_iface"]} -p tcp --dport {port} -j ACCEPT' for port in BOOTSTRAP_PORTS),
+            '# BOOTSTRAP RULES END',
+        ]
+
    if MANAGER.exists() and ctx['int_iface']:
        port = manager_port()
        lines += [
@@ -348,13 +372,16 @@ def manage_wireguard(ctx, changed):


 def manage_directory():
-    """Provision the user directory in slapd, whether or not a P2S system is configured."""
+    """Provision the user directory in slapd, whether or not a P2S system is configured; the problem, or ''."""
    if vpnrouter_ldap.ensure_services() or not unit_is_active('slapd'):
-        run(['systemctl', 'restart', 'slapd'])
+        result = run(['systemctl', 'restart', 'slapd'], check=False)
+        if result.returncode != 0:
+            return f'slapd cannot be restarted: {result.stderr.strip()}'
    try:
        vpnrouter_ldap.provision()
-    except ldap.LDAPError as exc:
-        fail(f'the user directory cannot be provisioned: {exc}')
+    except Exception as exc:
+        return f'the user directory cannot be provisioned: {exc!r}'
+    return ''


 def manage_openvpn(ctx, changed):
@@ -432,25 +459,62 @@ def refresh_web(cfg, ctx):
    }) + '\n', encoding='utf-8')


+def progress(state, step, error=None):
+    """Record where setup is in SETUP_STATE: running, deferred, done or failed, the step, and the error of a failure."""
+    SETUP_STATE.parent.mkdir(parents=True, exist_ok=True)
+    SETUP_STATE.write_text(json.dumps({
+        'state': state,
+        'step': step,
+        'error': error,
+        'time': time.strftime('%Y-%m-%dT%H:%M:%S%z'),
+    }) + '\n', encoding='utf-8')
+
+
+def start_bootstrap():
+    """Start the bootstrap services while no run has succeeded; a unit that is not installed is skipped."""
+    if not SETUP_DONE.exists():
+        for unit in BOOTSTRAP_UNITS:
+            run(['systemctl', 'start', '--no-block', unit], check=False)
+
+
+def end_bootstrap(ctx):
+    """After the first successful run: write SETUP_DONE, close the bootstrap ports and stop the bootstrap services."""
+    if SETUP_DONE.exists():
+        return
+    SETUP_DONE.parent.mkdir(parents=True, exist_ok=True)
+    SETUP_DONE.touch()
+    if setup_ufw(ctx):
+        run(['ufw', 'reload'], check=False)
+    for unit in BOOTSTRAP_UNITS:
+        run(['systemctl', 'stop', unit], check=False)
+
+
 def main():
+    start_bootstrap()
+    progress('running', 'configuration')
    cfg = vpnrouter.load_config()
    ctx, changed = vpnrouter.configure(cfg)

    if ctx is None:
        print('vpn-router: platform is none, configuration deferred; '
              f'set general.platform in {vpnrouter.CONFIG_FILE}')
+        progress('deferred', 'configuration', f'general.platform is none; set it in {vpnrouter.CONFIG_FILE}')
        return 0

+    progress('running', 'routes')
    if vpnrouter.routing_ready(ctx):
        apply_routes(ctx)

    if ctx['local_id_mode'] == 'public_ip' and ctx['local_fqdn']:
+        progress('running', 'public address')
        address = resolve_public_ip(ctx['local_fqdn'])
        ctx, more = vpnrouter.configure(cfg, local_id_override=address)
        changed |= more

+    progress('running', 'WireGuard keys')
    setup_wireguard(ctx)

+    progress('running', 'firewall')
    rules_changed = setup_ufw(ctx)
    if is_configured(ctx):
        status = run(['ufw', 'status'], check=False)
@@ -463,19 +527,53 @@ def main():
    elif rules_changed:
        run(['ufw', 'reload'], check=False)

-    manage_directory()
+    progress('running', 'strongSwan')
    manage_strongswan(ctx, changed)
+    progress('running', 'WireGuard')
    manage_wireguard(ctx, changed)
+
+    # The tunnels are up before the directory, whose problem ends setup only after every other step.
+    progress('running', 'user directory')
+    problem = manage_directory()
+    progress('running', 'OpenVPN')
    manage_openvpn(ctx, changed)

    if 'resolved' in changed:
+        progress('running', 'resolver')
        run(['systemctl', 'restart', 'systemd-resolved'], check=False)

+    progress('running', 'platform module')
    run_platform_setup(cfg, ctx)
+    progress('running', 'Docker rules')
    refresh_docker_rules()
+    progress('running', 'web service state')
    refresh_web(cfg, ctx)
+    if problem:
+        progress('failed', 'user directory', problem)
+        print(f'vpn-router: error: {problem}', file=sys.stderr)
+        return 1
+    progress('done', 'done')
+    end_bootstrap(ctx)
    return 0


+def record_failure(error):
+    """Mark the run failed in the step SETUP_STATE holds."""
+    try:
+        step = json.loads(SETUP_STATE.read_text(encoding='utf-8')).get('step')
+    except (OSError, ValueError):
+        step = None
+    progress('failed', step, error)
+
+
 if __name__ == '__main__':
-    sys.exit(main())
+    # A failing step ends with SystemExit carrying its message (vpnrouter.fail).
+    try:
+        code = main()
+    except SystemExit as exc:
+        record_failure(str(exc.code))
+        raise
+    except Exception as exc:
+        record_failure(repr(exc))
+        raise
+    sys.exit(code)
@@ -54,12 +54,9 @@ SCHEMA = {
    'wireguard':  {'enabled': False, 'address': '', 'listen_port': 51820},
    'openvpn':    {'enabled': False, 'address_pool': '', 'port': 1194, 'protocol': 'udp'},
    'docker':     {'mode': 'auto', 'exposed_ports': []},
-    'web':        {'enabled': True, 'port': 7081, 'jwt_secret': ''},
+    'web':        {'enabled': True, 'port': 7081, 'database': '', 'trusted_proxies': []},
 }

-#: The shortest secret that signs the tokens of the status web service.
-WEB_SECRET_MIN = 32
-
 DOCKER_MODES = ('auto', 'on', 'off')

 OPENVPN_PROTOCOLS = ('udp', 'tcp')
@@ -98,8 +95,8 @@ def warn(message):


 def fail(message):
-    print(f'vpn-router: error: {message}', file=sys.stderr)
-    sys.exit(1)
+    """End with status 1; SystemExit carries the message, which Python prints to standard error."""
+    sys.exit(f'vpn-router: error: {message}')


 def load_config(path=None):
@@ -262,9 +259,8 @@ def validate(cfg):
    if isinstance(web_port, bool) or not isinstance(web_port, int) or not 1 <= web_port <= 65535:
        raise ConfigError(f'web.port: {web_port!r} is not a valid port')

-    web_secret = _get(cfg, 'web', 'jwt_secret')
-    if web_secret and len(web_secret) < WEB_SECRET_MIN:
-        raise ConfigError(f'web.jwt_secret: use at least {WEB_SECRET_MIN} characters')
+    for item in _get_list(cfg, 'web', 'trusted_proxies'):
+        _check_address(item, 'web.trusted_proxies')

    docker_mode = _get(cfg, 'docker', 'mode')
    if docker_mode not in DOCKER_MODES:
@@ -0,0 +1,214 @@
+"""The tokens of the router's web service, for the web service and vpn-router-token.
+
+Tokens are HS256 JWTs signed with WEB_KEY, a key the web service makes on its first run on the
+router and that never leaves it. /auth issues an access token naming the user in sub, valid ACCESS_SECONDS, and
+a refresh token valid REFRESH_SECONDS; vpn-router-token issues tokens with the admin claim and
+any lifetime. A token without the admin claim must expire within MAX_LIFETIME seconds.
+
+psycopg is imported by driver() when the database is used, not when the module is loaded, so the
+module loads while libpq cannot be found yet. A driver that cannot be imported and every
+failure of the database raise DatabaseError, the only database exception callers see.
+
+Refresh tokens are random strings kept in PostgreSQL as their SHA-256, in the database of
+web.database, by default the vpn-router database on this host that provision() creates on the
+web service's first run, with the password in DATABASE_SECRET. A refresh token is used once: refreshing replaces it.
+"""
+
+import base64
+import contextlib
+import hashlib
+import hmac
+import json
+import pathlib
+import secrets
+import subprocess
+import time
+
+WEB_KEY = pathlib.Path('/etc/vpn-router/web.key')
+DATABASE_SECRET = pathlib.Path('/etc/vpn-router/database.secret')
+
+#: The role and the database provision() creates when web.database is not set.
+DATABASE = 'vpn-router'
+
+#: Seconds an access token of /auth and a refresh token are valid.
+ACCESS_SECONDS = 300
+REFRESH_SECONDS = 8 * 3600
+
+#: The longest a token without the admin claim may be valid, from now, in seconds.
+MAX_LIFETIME = 600
+
+#: Seconds of clock difference accepted between a token and this router.
+LEEWAY = 60
+
+SCHEMA = 'CREATE TABLE IF NOT EXISTS refresh_tokens (hash text PRIMARY KEY, name text NOT NULL, expires timestamptz NOT NULL)'
+
+
+class DatabaseError(Exception):
+    """The PostgreSQL driver cannot be loaded, or the database of the refresh tokens failed."""
+
+
+def key():
+    """The signing key, or '' while the web service has not made it."""
+    try:
+        return WEB_KEY.read_text(encoding='utf-8').strip()
+    except OSError:
+        return ''
+
+
+def ensure_key():
+    """Make the signing key when it is missing."""
+    if not key():
+        WEB_KEY.parent.mkdir(parents=True, exist_ok=True)
+        WEB_KEY.touch(mode=0o600)
+        WEB_KEY.chmod(0o600)
+        WEB_KEY.write_text(secrets.token_urlsafe(36) + '\n', encoding='utf-8')
+
+
+def b64encode(data):
+    return base64.urlsafe_b64encode(data).rstrip(b'=').decode()
+
+
+def b64decode(part):
+    return base64.urlsafe_b64decode(part + '=' * (-len(part) % 4))
+
+
+def sign(claims, secret):
+    head = b64encode(json.dumps({'alg': 'HS256', 'typ': 'JWT'}).encode())
+    body = b64encode(json.dumps(claims).encode())
+    signature = hmac.new(secret.encode(), f'{head}.{body}'.encode(), hashlib.sha256).digest()
+    return f'{head}.{body}.{b64encode(signature)}'
+
+
+def verify(token, secret):
+    """The claims of a token signed with HS256 by the secret, or None when it is not acceptable.
+
+    Only HS256 is accepted, so a token cannot choose a weaker algorithm. The token must expire
+    (exp); without the admin claim not later than MAX_LIFETIME seconds from now, so that a stolen
+    token is of little use. LEEWAY seconds of clock difference are allowed.
+    """
+    try:
+        header, payload, signature = token.split('.')
+        if json.loads(b64decode(header)).get('alg') != 'HS256':
+            return None
+        expected = hmac.new(secret.encode(), f'{header}.{payload}'.encode(), hashlib.sha256).digest()
+        if not hmac.compare_digest(expected, b64decode(signature)):
+            return None
+        claims = json.loads(b64decode(payload))
+        now = time.time()
+        expires = claims['exp']
+        if isinstance(expires, bool) or not isinstance(expires, (int, float)) or expires < now - LEEWAY:
+            return None
+        if claims.get('admin') is not True and expires > now + MAX_LIFETIME + LEEWAY:
+            return None
+        not_before = claims.get('nbf')
+        if not_before is not None and (isinstance(not_before, bool) or not isinstance(not_before, (int, float)) or not_before > now + LEEWAY):
+            return None
+    except (ValueError, KeyError, TypeError, AttributeError):
+        return None
+    return claims
+
+
+def admin_token(seconds):
+    """A token with the admin claim, valid for the seconds given."""
+    secret = key()
+    if not secret:
+        raise OSError(f'{WEB_KEY} does not exist; vpn-router-web makes it when it starts')
+    now = int(time.time())
+    return {'token': sign({'admin': True, 'iat': now, 'exp': now + seconds}, secret), 'expires': now + seconds}
+
+
+def driver():
+    """The psycopg module; DatabaseError while libpq cannot be found."""
+    try:
+        import psycopg
+    except ImportError as exc:
+        raise DatabaseError(f'the PostgreSQL driver cannot be loaded: {exc}') from exc
+    return psycopg
+
+
+def connection_string(cfg):
+    """web.database, or the database provision() creates; its password, of URL-safe characters, is written unquoted."""
+    configured = cfg['web']['database'].strip()
+    if configured:
+        return configured
+    password = DATABASE_SECRET.read_text(encoding='utf-8').strip()
+    return f'host=localhost dbname={DATABASE} user={DATABASE} password={password}'
+
+
+@contextlib.contextmanager
+def database(cfg):
+    """A connection to the database of the refresh tokens, with their table, closed at the end; DatabaseError on any failure."""
+    psycopg = driver()
+    try:
+        with psycopg.connect(connection_string(cfg), autocommit=True) as connection:
+            connection.execute(SCHEMA)
+            yield connection
+    except psycopg.Error as exc:
+        raise DatabaseError(f'the token database failed: {exc}') from exc
+
+
+def digest(token):
+    return hashlib.sha256(token.encode()).hexdigest()
+
+
+def issue(cfg, name):
+    """An access token and a refresh token for the user, as /auth answers them."""
+    secret = key()
+    now = int(time.time())
+    refresh = secrets.token_urlsafe(32)
+    with database(cfg) as connection:
+        connection.execute('DELETE FROM refresh_tokens WHERE expires < now()')
+        connection.execute("INSERT INTO refresh_tokens VALUES (%s, %s, now() + make_interval(secs => %s))",
+                           (digest(refresh), name, REFRESH_SECONDS))
+    return {
+        'access_token': sign({'sub': name, 'iat': now, 'exp': now + ACCESS_SECONDS}, secret),
+        'token_type': 'Bearer',
+        'expires_in': ACCESS_SECONDS,
+        'refresh_token': refresh,
+        'refresh_expires_in': REFRESH_SECONDS,
+    }
+
+
+def take(cfg, token):
+    """Remove an unexpired refresh token and return the user it was issued to, or None."""
+    if not isinstance(token, str) or not token:
+        return None
+    with database(cfg) as connection:
+        row = connection.execute('DELETE FROM refresh_tokens WHERE hash = %s RETURNING name, expires > now()',
+                                 (digest(token),)).fetchone()
+    return row[0] if row and row[1] else None
+
+
+def revoke(cfg, name):
+    """Remove all refresh tokens of the user."""
+    with database(cfg) as connection:
+        connection.execute('DELETE FROM refresh_tokens WHERE name = %s', (name,))
+
+
+def psql(sql, name):
+    """The output of an SQL statement, read from standard input, run by psql as postgres with name as the variable v1."""
+    args = ['runuser', '-u', 'postgres', '--', 'psql', '-X', '-q', '-t', '-A', '-v', 'ON_ERROR_STOP=1', '-v', f'v1={name}']
+    result = subprocess.run(args, input=sql, capture_output=True, text=True)
+    if result.returncode != 0:
+        raise OSError(f'psql failed: {result.stderr.strip()}')
+    return result.stdout.strip()
+
+
+def provision(cfg):
+    """Create the role, its password file and the database when web.database is not set and they are missing.
+
+    The password, of URL-safe characters, reaches psql on standard input, not on its command line.
+    """
+    if cfg['web']['database'].strip():
+        return
+    if not psql("SELECT 1 FROM pg_roles WHERE rolname = :'v1'", DATABASE):
+        DATABASE_SECRET.unlink(missing_ok=True)
+        psql("CREATE ROLE :\"v1\" LOGIN", DATABASE)
+    if not DATABASE_SECRET.is_file():
+        password = secrets.token_urlsafe(32)
+        psql(f"ALTER ROLE :\"v1\" PASSWORD '{password}'", DATABASE)
+        DATABASE_SECRET.touch(mode=0o600)
+        DATABASE_SECRET.chmod(0o600)
+        DATABASE_SECRET.write_text(password + '\n', encoding='utf-8')
+    if not psql("SELECT 1 FROM pg_database WHERE datname = :'v1'", DATABASE):
+        psql("CREATE DATABASE :\"v1\" OWNER :\"v1\"", DATABASE)
@@ -4,7 +4,13 @@ The package keeps its own database, dc=vpn-router,dc=local in /var/lib/ldap/vpn-
 next to the one slapd's installation creates. Root manages it through the local socket
 with SASL EXTERNAL: the database's root DN is root's peer-credential DN, so there is no
 administrator password. Users and groups are changed by the user-manager service account,
-which the ACLs allow to write ou=users and ou=groups and to set, never read, passwords.
+which the ACLs allow to write ou=users and ou=groups and to set, never read, passwords and
+TOTP secrets.
+
+The password policy locks an account for LOCKOUT_SECONDS after LOCKOUT_FAILURES failed binds;
+a successful bind clears the count. slapd's otp overlay is configured on the database, so a
+bind of a user with a TOTP token takes the password followed by a current code.
+cn=totp,ou=policies holds the TOTP parameters the token of every user points to.
 """

 import os
@@ -24,6 +30,7 @@ GROUPS = f'ou=groups,{BASE}'
 SERVICES = f'ou=services,{BASE}'
 POLICIES = f'ou=policies,{BASE}'
 DEFAULT_POLICY = f'cn=default,{POLICIES}'
+TOTP_PARAMS = f'cn=totp,{POLICIES}'
 MANAGER = f'cn=user-manager,{SERVICES}'
 ROOT = 'gidNumber=0+uidNumber=0,cn=peercred,cn=external,cn=auth'

@@ -33,12 +40,26 @@ MANAGER_SECRET = SECRET_DIR / 'user-manager.secret'
 SLAPD_DEFAULTS = pathlib.Path('/etc/default/slapd')
 SLAPD_SERVICES = 'SLAPD_SERVICES="ldapi:///"'

-MODULES = ('argon2', 'ppolicy')
+MODULES = ('argon2', 'ppolicy', 'otp')
 PASSWORD_HASH = '{ARGON2}'

+#: Failed binds that lock an account, within LOCKOUT_SECONDS, and the seconds it stays locked.
+LOCKOUT_FAILURES = 5
+LOCKOUT_SECONDS = 900
+
+#: The overlays of the database, in the order they are configured.
+OVERLAYS = ('ppolicy', 'otp')
+
+#: The TOTP parameters: digits, the OID of HMAC-SHA1, seconds per time step, and the time steps accepted before and after now.
+TOTP_DIGITS = 6
+TOTP_ALGORITHM = '1.2.840.113549.2.7'
+TOTP_PERIOD = 30
+TOTP_WINDOW = 1
+
 #: The ACLs of the database, in order.
 ACCESS = (
    f'to attrs=userPassword by dn.exact="{MANAGER}" =wx by self =wx by anonymous auth by * none',
+    f'to attrs=oathSecret by dn.exact="{MANAGER}" =w by * none',
    f'to dn.subtree="{USERS}" by dn.exact="{MANAGER}" manage by * read',
    f'to dn.subtree="{GROUPS}" by dn.exact="{MANAGER}" manage by * read',
    'to * by * read',
@@ -118,13 +139,18 @@ def _ensure_database(connection):
    return dn


-def _ensure_overlay(connection, database):
-    if not connection.search_s(database, ldap.SCOPE_ONELEVEL, '(objectClass=olcPPolicyConfig)', ['1.1']):
-        connection.add_s(f'olcOverlay=ppolicy,{database}', ldap.modlist.addModlist({
-            'objectClass': [b'olcOverlayConfig', b'olcPPolicyConfig'],
-            'olcOverlay': [b'ppolicy'],
-            'olcPPolicyDefault': [DEFAULT_POLICY.encode()],
-        }))
+def _ensure_overlays(connection, database):
+    """Add the overlays of OVERLAYS the database lacks; the otp overlay has no settings of its own."""
+    found = connection.search_s(database, ldap.SCOPE_ONELEVEL, '(objectClass=olcOverlayConfig)', ['olcOverlay'])
+    present = {ORDER.sub(b'', attributes['olcOverlay'][0]).decode() for _, attributes in found}
+    for name in OVERLAYS:
+        if name in present:
+            continue
+        record = {'objectClass': [b'olcOverlayConfig'], 'olcOverlay': [name.encode()]}
+        if name == 'ppolicy':
+            record = {'objectClass': [b'olcOverlayConfig', b'olcPPolicyConfig'], 'olcOverlay': [b'ppolicy'],
+                      'olcPPolicyDefault': [DEFAULT_POLICY.encode()]}
+        connection.add_s(f'olcOverlay={name},{database}', ldap.modlist.addModlist(record))


 def _ensure_entry(connection, dn, attributes):
@@ -168,7 +194,7 @@ def provision():
        _ensure_modules(connection)
        _ensure_password_hash(connection)
        database = _ensure_database(connection)
-        _ensure_overlay(connection, database)
+        _ensure_overlays(connection, database)
        _ensure_entry(connection, BASE, {'objectClass': [b'dcObject', b'organization'], 'dc': [b'vpn-router'], 'o': [b'vpn-router']})
        for unit in (USERS, GROUPS, SERVICES, POLICIES):
            _ensure_entry(connection, unit, {'objectClass': [b'organizationalUnit'], 'ou': [unit.split(',')[0][3:].encode()]})
@@ -177,7 +203,17 @@ def provision():
            'cn': [b'default'],
            'pwdAttribute': [b'userPassword'],
            'pwdLockout': [b'TRUE'],
-            'pwdMaxFailure': [b'0'],
+            'pwdMaxFailure': [str(LOCKOUT_FAILURES).encode()],
+            'pwdLockoutDuration': [str(LOCKOUT_SECONDS).encode()],
+            'pwdFailureCountInterval': [str(LOCKOUT_SECONDS).encode()],
+        })
+        _ensure_entry(connection, TOTP_PARAMS, {
+            'objectClass': [b'organizationalRole', b'oathTOTPParams'],
+            'cn': [b'totp'],
+            'oathOTPLength': [str(TOTP_DIGITS).encode()],
+            'oathHMACAlgorithm': [TOTP_ALGORITHM.encode()],
+            'oathTOTPTimeStepPeriod': [str(TOTP_PERIOD).encode()],
+            'oathTOTPTimeStepWindow': [str(TOTP_WINDOW).encode()],
        })
        _ensure_manager(connection)
    finally:
@@ -1,6 +1,7 @@
 """Interfaces and addresses of the router, read from the system.

-Only the Python standard library is used, so the web service can use this module too.
+Only the Python standard library is used, so the web service and its bootstrap server can use
+this module too.
 """

 import pathlib
@@ -107,3 +108,28 @@ def internal_address(mode, internal, int_addr):
    except ConfigError:
        return ''
    return addresses[0] if addresses else ''
+
+
+def web_listen(configuration, default_port):
+    """Where the web service listens, from vpn-router.json read as configuration, as ((address, port), '').
+
+    The address is the internal address as general.mode directs, the port web.port, else
+    default_port. While web.enabled is false or the address is not known, the answer is (None,
+    the reason).
+    """
+    def section(name):
+        value = configuration.get(name)
+        return value if isinstance(value, dict) else {}
+
+    web = section('web')
+    if web.get('enabled', True) is False:
+        return None, 'the web service is disabled in the configuration'
+    port = web.get('port', default_port)
+    if isinstance(port, bool) or not isinstance(port, int) or not 1 <= port <= 65535:
+        port = default_port
+    general, interfaces, local = section('general'), section('interfaces'), section('local')
+    address = internal_address(str(general.get('mode') or 'manual'), str(interfaces.get('internal') or ''),
+                               str(local.get('int_addr') or ''))
+    if not address:
+        return None, 'the internal address is not known yet'
+    return (address, port), ''
@@ -0,0 +1,40 @@
+"""The listening loop of the web service and of its bootstrap server.
+
+Only the Python standard library is used, so the bootstrap server, which must run whatever else
+is broken, can use it.
+"""
+
+import threading
+import time
+from http.server import ThreadingHTTPServer
+
+
+def serve(wanted, handler, log, interval):
+    """Work out where to listen every interval seconds with wanted, and run an HTTP server of handler there.
+
+    wanted answers ((address, port), '') or (None, the reason not to listen). When the address or
+    the port changes, the server is stopped and a new one listens at the new place; while there is
+    no place or it cannot be bound, no server runs and the next round tries again. Every new
+    reason is logged with log. Does not return.
+    """
+    server, thread, listening, reported = None, None, None, None
+    while True:
+        target, reason = wanted()
+        if target != listening and server is not None:
+            server.shutdown()
+            server.server_close()
+            thread.join()
+            server, listening = None, None
+        if target is not None and server is None:
+            try:
+                server = ThreadingHTTPServer(target, handler)
+            except OSError as exc:
+                reason = f'cannot listen on {target[0]}:{target[1]}: {exc}'
+            else:
+                thread = threading.Thread(target=server.serve_forever, name='server', daemon=True)
+                thread.start()
+                listening, reason = target, f'serving on http://{target[0]}:{target[1]}'
+        if reason and reason != reported:
+            log(reason)
+            reported = reason
+        time.sleep(interval)
@@ -10,24 +10,41 @@ optional description and at least one member; a group cannot be empty, so the la
 cannot leave it and a user who is the last member of a group cannot be removed. Groups only
 record membership; nothing on the router reads them.

+A user may have a TOTP token: the user entry itself is an oathTOTPUser and oathTOTPToken whose
+oathTOTPToken points to itself, with the secret in oathSecret and the parameters of
+vpnrouter_ldap.TOTP_PARAMS. Enrolment stores nothing until a code of the new secret is
+verified: otp_new makes a secret and its otpauth URI, and otp_enrol stores the secret given a
+code of it. slapd's otp overlay then takes the password followed by a code at every bind.
+
+A user is disabled when pwdAccountLockedTime holds LOCKED, and locked when the password
+policy has set it after failed binds, for vpnrouter_ldap.LOCKOUT_SECONDS; unlock ends that.
+login() checks a user's password and code with a bind. Members of ADMIN_GROUP manage the
+router through the web service. Changing the password, disabling, unlocking and removing a
+user revoke its refresh tokens.
+
 Every operation runs in a session(), which holds the lock that serializes changes and the
 connection of the user manager account. A request that cannot be done raises Failure, whose
-status says why: 2 the request is invalid, 3 the user does not exist, 4 the user exists, 1
-anything else. Answers are dicts with the keys of the JSON answers.
+status says why: 2 the request is invalid, 3 the user does not exist, 4 the user exists, 5 a
+login failed, 1 anything else. Answers are dicts with the keys of the JSON answers.
 """

+import base64
+import calendar
 import contextlib
 import fcntl
 import pathlib
 import re
 import subprocess
 import sys
+import time

 import ldap
 import ldap.filter
 import ldap.modlist
+import pyotp

 import vpnrouter
+import vpnrouter_auth
 import vpnrouter_ldap

 #: A user name is one lowercase DNS label, as it is also the common name of the certificate.
@@ -42,18 +59,27 @@ EMAIL_MAX = 254
 #: The keys of the person attributes and their LDAP attributes.
 PERSON = (('first_name', 'givenName'), ('last_name', 'sn'), ('email', 'mail'))
 #: The attributes read for a user.
-ATTRIBUTES = ['uid', 'pwdAccountLockedTime', 'givenName', 'sn', 'mail']
+ATTRIBUTES = ['uid', 'pwdAccountLockedTime', 'givenName', 'sn', 'mail', 'oathTOTPToken']
+#: The readable attributes of a TOTP token; oathSecret is never readable.
+TOKEN_ATTRIBUTES = ['oathTOTPToken', 'oathTOTPParams', 'oathSecretTime', 'oathTOTPLastTimeStep', 'oathTOTPTimeStepDrift',
+                    'oathLastLogin', 'oathLastFailure', 'oathFailureCount']
+#: A TOTP secret as otp_new makes it: 32 base32 characters, 160 bits.
+TOTP_SECRET = re.compile(r'^[A-Z2-7]{32}$')
 #: The attributes read for a group.
 GROUP_ATTRIBUTES = ['cn', 'description', 'member']
 #: The longest group description, in characters.
 DESCRIPTION_MAX = 256
-#: The value of pwdAccountLockedTime that locks an account until it is unlocked.
+#: The value of pwdAccountLockedTime that disables an account until it is enabled.
 LOCKED = b'000001010000Z'
+#: The group whose members manage the router through the web service.
+ADMIN_GROUP = 'admins'
+#: The digits of a TOTP code given to login().
+CODE = re.compile(r'^\d{6}$')
 LOCK_FILE = pathlib.Path('/run/vpn-router/users.lock')


 class Failure(Exception):
-    """A request that cannot be done; status is 2 invalid, 3 no such user, 4 the user exists, 1 other."""
+    """A request that cannot be done; status is 2 invalid, 3 no such user, 4 the user exists, 5 failed login, 1 other."""

    def __init__(self, message, status=1):
        super().__init__(message)
@@ -138,13 +164,24 @@ def publish_crl():
        subprocess.run(['swanctl', '--load-creds'], capture_output=True, text=True)


+def locked_since(value):
+    """The time a pwdAccountLockedTime value names, in seconds since the epoch, or 0 when it names none."""
+    try:
+        return calendar.timegm(time.strptime(value[:14].decode(), '%Y%m%d%H%M%S'))
+    except ValueError:
+        return 0
+
+
 def summary(attributes):
    """A user as the answers show it; a person attribute the entry does not have is None."""
    name = attributes['uid'][0].decode()
+    lock = attributes.get('pwdAccountLockedTime', [None])[0]
    user = {
        'name': name,
-        'enabled': 'pwdAccountLockedTime' not in attributes,
+        'enabled': lock != LOCKED,
+        'locked': lock not in (None, LOCKED) and locked_since(lock) > time.time() - vpnrouter_ldap.LOCKOUT_SECONDS,
        'certificate': certificate(name)[0].is_file(),
+        'otp': 'oathTOTPToken' in attributes,
    }
    for key, attribute in PERSON:
        values = attributes.get(attribute)
@@ -203,21 +240,107 @@ def passwd(connection, name, password):
    check_password(password)
    entry(connection, name)
    connection.passwd_s(dn_of(name), None, password)
+    vpnrouter_auth.revoke(vpnrouter.load_config(), name)
    return summary(entry(connection, name))


 def disable(connection, name):
-    if 'pwdAccountLockedTime' not in entry(connection, name):
+    if entry(connection, name).get('pwdAccountLockedTime') != [LOCKED]:
        connection.modify_s(dn_of(name), [(ldap.MOD_REPLACE, 'pwdAccountLockedTime', [LOCKED])])
+    vpnrouter_auth.revoke(vpnrouter.load_config(), name)
    return summary(entry(connection, name))


 def enable(connection, name):
-    if 'pwdAccountLockedTime' in entry(connection, name):
+    if entry(connection, name).get('pwdAccountLockedTime') == [LOCKED]:
        connection.modify_s(dn_of(name), [(ldap.MOD_DELETE, 'pwdAccountLockedTime', None)])
    return summary(entry(connection, name))


+def unlock(connection, name):
+    """End a lockout after failed binds and clear the failure count; a disabled user stays disabled."""
+    entry(connection, name)
+    current = connection.search_s(dn_of(name), ldap.SCOPE_BASE, attrlist=['pwdAccountLockedTime', 'pwdFailureTime'])[0][1]
+    changes = [(ldap.MOD_DELETE, 'pwdFailureTime', None)] if 'pwdFailureTime' in current else []
+    if 'pwdAccountLockedTime' in current and current['pwdAccountLockedTime'] != [LOCKED]:
+        changes.append((ldap.MOD_DELETE, 'pwdAccountLockedTime', None))
+    if changes:
+        connection.modify_s(dn_of(name), changes)
+    vpnrouter_auth.revoke(vpnrouter.load_config(), name)
+    return summary(entry(connection, name))
+
+
+def login(name, password, code):
+    """Check the user's password, and the TOTP code of an enrolled user, with a bind as the user.
+
+    A wrong name, password or code, a locked or a disabled user give the same Failure, status 5.
+    """
+    refused = Failure('the name, the password or the code is wrong', 5)
+    if not isinstance(name, str) or not NAME.match(name) or not isinstance(password, str) or not password:
+        raise refused
+    if code is not None and code != '' and (not isinstance(code, str) or not CODE.match(code)):
+        raise refused
+    connection = ldap.initialize(vpnrouter_ldap.URI)
+    try:
+        connection.simple_bind_s(dn_of(name), password + (code or ''))
+    except ldap.INVALID_CREDENTIALS:
+        raise refused
+    finally:
+        connection.unbind_s()
+
+
+def is_admin(connection, name):
+    """Whether the user exists and is a member of ADMIN_GROUP."""
+    try:
+        entry(connection, name)
+    except Failure:
+        return False
+    return ADMIN_GROUP in groups_of(connection, name)
+
+
+def totp(secret):
+    return pyotp.TOTP(secret, digits=vpnrouter_ldap.TOTP_DIGITS, interval=vpnrouter_ldap.TOTP_PERIOD)
+
+
+def otp_new(connection, name):
+    """A new TOTP secret for the user and its otpauth URI, issued by the router's wan.local_fqdn; nothing is stored."""
+    entry(connection, name)
+    secret = pyotp.random_base32()
+    issuer = vpnrouter.load_config()['wan']['local_fqdn'].strip()
+    return {'name': name, 'secret': secret, 'uri': totp(secret).provisioning_uri(name=name, issuer_name=issuer or None)}
+
+
+def otp_enrol(connection, name, secret, code):
+    """Store the secret of otp_new as the user's TOTP token, replacing one the user has, when code is a code of it."""
+    current = entry(connection, name)
+    if not isinstance(secret, str) or not TOTP_SECRET.match(secret):
+        raise Failure('the secret must be 32 base32 characters, as a new secret has them', 2)
+    if not isinstance(code, str) or not re.fullmatch(rf'\d{{{vpnrouter_ldap.TOTP_DIGITS}}}', code):
+        raise Failure(f'the code must be {vpnrouter_ldap.TOTP_DIGITS} digits', 2)
+    if not totp(secret).verify(code, valid_window=vpnrouter_ldap.TOTP_WINDOW):
+        raise Failure('the code does not match the secret; check the clock of the authenticator', 2)
+    changes = [] if 'oathTOTPToken' in current else [(ldap.MOD_ADD, 'objectClass', [b'oathTOTPUser', b'oathTOTPToken'])]
+    changes += [
+        (ldap.MOD_REPLACE, 'oathTOTPToken', [dn_of(name).encode()]),
+        (ldap.MOD_REPLACE, 'oathTOTPParams', [vpnrouter_ldap.TOTP_PARAMS.encode()]),
+        (ldap.MOD_REPLACE, 'oathSecret', [base64.b32decode(secret)]),
+        (ldap.MOD_REPLACE, 'oathSecretTime', [time.strftime('%Y%m%d%H%M%SZ', time.gmtime()).encode()]),
+    ]
+    connection.modify_s(dn_of(name), changes)
+    return summary(entry(connection, name))
+
+
+def otp_remove(connection, name):
+    """Remove the user's TOTP token, when the user has one."""
+    if 'oathTOTPToken' in entry(connection, name):
+        token = connection.search_s(dn_of(name), ldap.SCOPE_BASE, attrlist=TOKEN_ATTRIBUTES)[0][1]
+        changes = [(ldap.MOD_DELETE, 'objectClass', [b'oathTOTPUser', b'oathTOTPToken']),
+                   (ldap.MOD_DELETE, 'oathSecret', None)]
+        changes += [(ldap.MOD_DELETE, attribute, None) for attribute in token]
+        connection.modify_s(dn_of(name), changes)
+    return summary(entry(connection, name))
+
+
 def remove(connection, name):
    """Remove the user from its groups and the directory, and revoke the certificate.

@@ -238,6 +361,7 @@ def remove(connection, name):
    cert.unlink(missing_ok=True)
    key.unlink(missing_ok=True)
    connection.delete_s(dn_of(name))
+    vpnrouter_auth.revoke(vpnrouter.load_config(), name)
    return {'name': name, 'removed': True}


@@ -11,32 +11,42 @@ Pages:
  /config  the configuration without its secrets, and what the router resolved and
           runs, as JSON
  /users   the users in the router's directory, whether or not a P2S system is enabled: list,
-           add, change, set the password, disable, enable, remove, and the OpenVPN client
-           profile (see users())
+           add, change, set the password, disable, enable, remove, the OpenVPN client
+           profile, and TOTP enrolment (see users())
  /groups  the groups of the router's directory: list, add, rename or describe, remove, and
           add and take out members (see groups())
+  /auth    sign-in with a user's password and TOTP code, refresh and sign-out (see auth())

-/config, /users and /groups need an HS256 JWT in an Authorization: Bearer header that is signed
-with web.jwt_secret and expires within ten minutes; they are not served (404) while
-web.jwt_secret is not set.
+/config, /users and /groups need an HS256 JWT in an Authorization: Bearer header signed with
+vpnrouter_auth.WEB_KEY (see vpnrouter_auth.verify). A token of /auth names a user in sub, who
+must be a member of vpnrouter_users.ADMIN_GROUP; a token of vpn-router-token has the admin
+claim, and only such a token may change ADMIN_GROUP: create, rename to it, change, remove it,
+add or take out its members, and remove one of its members. These paths and /auth are not
+served (404) while the key does not exist. /auth is limited to AUTH_LIMIT calls per client
+address in AUTH_WINDOW seconds; the address of a call from one of web.trusted_proxies, by
+default the address the service listens on, is the last entry of its X-Forwarded-For header.

 / and /status show only the status: whether the services run, the package version, whether
 Docker is installed, and the state of site-to-site, road-warrior (IPsec P2S), OpenVPN and
 WireGuard. No configuration data, key or address is shown.

-The service needs nothing else to start and does not end on its own. A discovery thread works
-out the internal address from /etc/vpn-router/vpn-router.json as general.mode directs, with
-web.port and web.enabled, every DISCOVERY_SECONDS, and runs the server thread there: it starts
-it, moves it when the address or the port changes, and stops it while the service is disabled
-or the address is not known. Every request reads what it needs when it is made: the token secret
-from the configuration, and the features of the router and the facts shown by /config from
+The service first initializes, without listening: it loads the PostgreSQL driver
+(vpnrouter_auth.driver; psycopg finds libpq through the linker cache, which the package manager
+refreshes only at the end of an installation), then it makes the signing key and the refresh
+tokens' database (vpnrouter_auth), retrying every INIT_RETRY_SECONDS until both are ready and
+logging each failure. Then it listens.
+
+The service needs nothing else to start and does not end on its own. Every DISCOVERY_SECONDS it
+works out the internal address from /etc/vpn-router/vpn-router.json as general.mode directs,
+with web.port and web.enabled (vpnrouter_net.web_listen), and runs the server there
+(vpnrouter_serve.serve): it starts it, moves it when the address or the port changes, and stops
+it while the service is disabled or the address is not known. Every request reads what it needs when it is made: the signing
+key, the configuration, and the features of the router and the facts shown by /config from
 /run/vpn-router/web.json, which vpn-router-setup writes. Before setup has written that file
 the router has no features, and /config shows no configuration or interfaces.
 """

-import base64
-import hashlib
-import hmac
+import collections
 import json
 import os
 import signal
@@ -45,18 +55,24 @@ import sys
 import threading
 import time
 from html import escape
-from http.server import BaseHTTPRequestHandler, ThreadingHTTPServer
+from http.server import BaseHTTPRequestHandler

 sys.path.insert(0, os.path.dirname(os.path.realpath(__file__)))

 import ldap

+import vpnrouter
+import vpnrouter_auth
 import vpnrouter_net
+import vpnrouter_serve
 import vpnrouter_users

 STATE_FILE = '/run/vpn-router/web.json'
 CONFIG_FILE = '/etc/vpn-router/vpn-router.json'

+#: Seconds between two attempts of the initialization.
+INIT_RETRY_SECONDS = 15
+
 #: The port when web.port is not set.
 DEFAULT_PORT = 7081

@@ -65,11 +81,9 @@ DISCOVERY_SECONDS = 30
 COMMAND_TIMEOUT = 5
 CACHE_SECONDS = 2

-#: Seconds of clock difference accepted between a token and this router.
-LEEWAY = 60
-
-#: The longest a token may be valid, from now, in seconds.
-MAX_LIFETIME = 600
+#: The calls of /auth allowed per client address in AUTH_WINDOW seconds.
+AUTH_LIMIT = 10
+AUTH_WINDOW = 60

 TAILWIND = 'https://cdn.jsdelivr.net/npm/@tailwindcss/browser@4'

@@ -82,11 +96,15 @@ OPENVPN_STATUS = '/run/openvpn-server/status-vpn-router.log'
 BODY_LIMIT = 4096

 #: The HTTP status for each status of a vpnrouter_users.Failure; any other is 500.
-USER_STATUS = {2: 400, 3: 404, 4: 409}
+USER_STATUS = {2: 400, 3: 404, 4: 409, 5: 401}

 _cache = {'at': 0.0, 'data': None}
 _lock = threading.Lock()

+#: The times of the recent /auth calls of each client address.
+_auth_calls = collections.defaultdict(collections.deque)
+_auth_lock = threading.Lock()
+

 def read_json(path):
    """The object in a JSON file, or {} when the file is missing or holds no object."""
@@ -105,14 +123,12 @@ def section(configuration, name):


 def current():
-    """What a request needs, read now: the features and facts setup resolved, and the token secret.
+    """What a request needs, read now: the features and facts setup resolved, and the signing key.

-    Before setup has run the router has no features, and without a secret /config and /users are
-    not served.
+    Before setup has run the router has no features, and without a key the protected paths and
+    /auth are not served.
    """
-    state = read_json(STATE_FILE)
-    secret = section(read_json(CONFIG_FILE), 'web').get('jwt_secret')
-    return dict(state, jwt_secret=secret if isinstance(secret, str) else '')
+    return dict(read_json(STATE_FILE), key=vpnrouter_auth.key())


 def run(*args):
@@ -253,37 +269,6 @@ def collect(state):
    return status


-def b64decode(part):
-    return base64.urlsafe_b64decode(part + '=' * (-len(part) % 4))
-
-
-def verify(token, secret):
-    """The claims of a token signed with HS256 by the secret, or None when it is not acceptable.
-
-    Only HS256 is accepted, so a token cannot choose a weaker algorithm. The token
-    must expire (exp), and not later than MAX_LIFETIME seconds from now, so that a
-    stolen token is of little use; LEEWAY seconds of clock difference are allowed.
-    """
-    try:
-        header, payload, signature = token.split('.')
-        if json.loads(b64decode(header)).get('alg') != 'HS256':
-            return None
-        expected = hmac.new(secret.encode(), f'{header}.{payload}'.encode(), hashlib.sha256).digest()
-        if not hmac.compare_digest(expected, b64decode(signature)):
-            return None
-        claims = json.loads(b64decode(payload))
-        now = time.time()
-        expires = claims['exp']
-        if isinstance(expires, bool) or not isinstance(expires, (int, float)) or not now - LEEWAY <= expires <= now + MAX_LIFETIME + LEEWAY:
-            return None
-        not_before = claims.get('nbf')
-        if not_before is not None and (isinstance(not_before, bool) or not isinstance(not_before, (int, float)) or not_before > now + LEEWAY):
-            return None
-    except (ValueError, KeyError, TypeError, AttributeError):
-        return None
-    return claims
-
-
 def lines_of(*args):
    """The output lines of a command, or None when it fails."""
    result = run(*args)
@@ -420,8 +405,12 @@ def users(state, method, path, body):
      PUT    /users/<name>/password   set the password: {"password": ...}
      POST   /users/<name>/disable    lock the user out of OpenVPN
      POST   /users/<name>/enable     let the user in again
+      POST   /users/<name>/unlock     end a lockout after failed logins
      DELETE /users/<name>            remove the user and revoke the certificate
      GET    /users/<name>/profile    the OpenVPN client profile
+      POST   /users/<name>/otp        a new TOTP secret and its otpauth URI; nothing is stored
+      PUT    /users/<name>/otp        store a TOTP secret: {"secret", "code"}, code a code of the secret
+      DELETE /users/<name>/otp        remove the TOTP token
    """
    if body == 'invalid' or (body is not None and not isinstance(body, dict)):
        return 400, {'error': 'the body must be a JSON object'}
@@ -436,8 +425,12 @@ def users(state, method, path, body):
        ('PUT', 3, 'password'): lambda connection: vpnrouter_users.passwd(connection, name, body.get('password')),
        ('POST', 3, 'disable'): lambda connection: vpnrouter_users.disable(connection, name),
        ('POST', 3, 'enable'): lambda connection: vpnrouter_users.enable(connection, name),
+        ('POST', 3, 'unlock'): lambda connection: vpnrouter_users.unlock(connection, name),
        ('DELETE', 2, None): lambda connection: vpnrouter_users.remove(connection, name),
        ('GET', 3, 'profile'): lambda connection: vpnrouter_users.profile(connection, name),
+        ('POST', 3, 'otp'): lambda connection: vpnrouter_users.otp_new(connection, name),
+        ('PUT', 3, 'otp'): lambda connection: vpnrouter_users.otp_enrol(connection, name, body.get('secret'), body.get('code')),
+        ('DELETE', 3, 'otp'): lambda connection: vpnrouter_users.otp_remove(connection, name),
    }
    if route not in actions:
        return 404, {'error': f'no route {method} {path}'}
@@ -448,9 +441,11 @@ def users(state, method, path, body):
        return USER_STATUS.get(failure.status, 500), {'error': str(failure)}
    except (ldap.LDAPError, OSError) as exc:
        return 500, {'error': f'the user directory failed: {exc}'}
-    except SystemExit:
-        # vpnrouter ends with SystemExit on a configuration it cannot read; it has printed the reason.
-        return 500, {'error': 'the router configuration cannot be read'}
+    except vpnrouter_auth.DatabaseError as exc:
+        return 500, {'error': str(exc)}
+    except SystemExit as exc:
+        # vpnrouter ends with SystemExit, carrying the reason, on a configuration it cannot read.
+        return 500, {'error': str(exc.code)}
    return (201 if route == ('POST', 1, None) else 200), document


@@ -492,6 +487,85 @@ def groups(state, method, path, body):
    return (201 if route == ('POST', 1, None) else 200), document


+def changes_admins(connection, method, parts, body):
+    """Whether a /users or /groups request creates, renames to, changes or removes ADMIN_GROUP, or removes a member of it."""
+    admins = vpnrouter_users.ADMIN_GROUP
+    if parts[0] == 'groups':
+        if method == 'POST' and len(parts) == 1:
+            return body.get('name') == admins
+        if method != 'GET' and len(parts) > 1 and parts[1] == admins:
+            return True
+        return method == 'PUT' and len(parts) == 2 and body.get('name') == admins
+    if parts[0] == 'users' and method == 'DELETE' and len(parts) == 2:
+        return admins in vpnrouter_users.groups_of(connection, parts[1])
+    return False
+
+
+def refusal(claims, method, path, body):
+    """Why the holder of the claims may not make the request to a protected path, or '' when it may."""
+    if claims.get('admin') is True:
+        return ''
+    name = claims.get('sub')
+    if not isinstance(name, str):
+        return 'the token names no user'
+    admins = vpnrouter_users.ADMIN_GROUP
+    with vpnrouter_users.session() as connection:
+        if not vpnrouter_users.is_admin(connection, name):
+            return f'user {name} is not a member of {admins}'
+        if changes_admins(connection, method, path.strip('/').split('/'), body if isinstance(body, dict) else {}):
+            return f'only a token of vpn-router-token may change the group {admins} or remove its members'
+    return ''
+
+
+def auth(path, body):
+    """Answer a POST to /auth with vpnrouter_users and vpnrouter_auth: the HTTP status and the JSON document.
+
+      POST /auth           sign in: {"name", "password", "code"?}; the tokens
+      POST /auth/refresh   {"refresh_token"}; new tokens, the refresh token used up
+      POST /auth/logout    {"refresh_token"}; the refresh token revoked
+    """
+    if not isinstance(body, dict):
+        return 400, {'error': 'the body must be a JSON object'}
+    try:
+        cfg = vpnrouter.load_config()
+        if path == '/auth':
+            vpnrouter_users.login(body.get('name'), body.get('password'), body.get('code'))
+            return 200, vpnrouter_auth.issue(cfg, body['name'])
+        name = vpnrouter_auth.take(cfg, body.get('refresh_token'))
+        if path == '/auth/logout':
+            return 200, {'revoked': name is not None}
+        if name is None:
+            return 401, {'error': 'the refresh token is not valid'}
+        with vpnrouter_users.session() as connection:
+            user = vpnrouter_users.summary(vpnrouter_users.entry(connection, name))
+        if not user['enabled'] or user['locked']:
+            return 401, {'error': f'user {name} is disabled or locked'}
+        return 200, vpnrouter_auth.issue(cfg, name)
+    except vpnrouter_users.Failure as failure:
+        return (401 if failure.status == 3 else USER_STATUS.get(failure.status, 500)), {'error': str(failure)}
+    except (ldap.LDAPError, OSError) as exc:
+        return 500, {'error': f'the user directory failed: {exc}'}
+    except vpnrouter_auth.DatabaseError as exc:
+        return 500, {'error': str(exc)}
+    except SystemExit as exc:
+        return 500, {'error': str(exc.code)}
+
+
+def auth_wait(address):
+    """Seconds until the client address may call /auth again, 0 when it may now; a call that may is counted."""
+    now = time.monotonic()
+    with _auth_lock:
+        for idle in [key for key, times in _auth_calls.items() if times[-1] <= now - AUTH_WINDOW]:
+            del _auth_calls[idle]
+        calls = _auth_calls[address]
+        while calls and calls[0] <= now - AUTH_WINDOW:
+            calls.popleft()
+        if len(calls) >= AUTH_LIMIT:
+            return int(calls[0] + AUTH_WINDOW - now) + 1
+        calls.append(now)
+        return 0
+
+
 class Handler(BaseHTTPRequestHandler):
    protocol_version = 'HTTP/1.1'

@@ -502,29 +576,65 @@ class Handler(BaseHTTPRequestHandler):
                _cache['at'] = time.monotonic()
            return _cache['data']

-    def authorized(self, secret):
-        """Whether the request carries a token signed with the secret; False also while no secret is set."""
+    def claims(self, secret):
+        """The claims of the request's token signed with the secret, or None without an acceptable one."""
        header = self.headers.get('Authorization', '')
-        return bool(secret) and header.startswith('Bearer ') and verify(header[7:].strip(), secret) is not None
+        return vpnrouter_auth.verify(header[7:].strip(), secret) if header.startswith('Bearer ') else None
+
+    def client(self):
+        """The client address: the last X-Forwarded-For entry when the peer is a trusted proxy, else the peer."""
+        peer = self.client_address[0]
+        trusted = section(read_json(CONFIG_FILE), 'web').get('trusted_proxies') or [self.server.server_address[0]]
+        forwarded = self.headers.get('X-Forwarded-For', '')
+        if isinstance(trusted, list) and peer in trusted and forwarded.strip():
+            return forwarded.split(',')[-1].strip()
+        return peer
+
+    def refusal(self, claims, path, payload):
+        """The HTTP status and the message that refuse a request to a protected path, or None when it may be made."""
+        try:
+            reason = refusal(claims, self.command, path, payload)
+        except (ldap.LDAPError, OSError) as exc:
+            return 500, f'the user directory failed: {exc}'
+        return (403, reason) if reason else None

    def answer(self, send_body):
        path = self.path.split('?', 1)[0]
        extra = []
+        self.principal = None
        # A body that is not read must not be taken for the next request.
        if self.command not in ('GET', 'HEAD'):
            self.close_connection = True
        directory = next((name for name in ('users', 'groups') if path == f'/{name}' or path.startswith(f'/{name}/')), None)
        protected = path == '/config' or directory is not None
-        state = current() if protected else {}
-        if protected and not state['jwt_secret']:
+        signing = path in ('/auth', '/auth/refresh', '/auth/logout')
+        state = current() if protected or signing else {}
+        claims = self.claims(state['key']) if protected and state['key'] else None
+        payload = self.request_body() if (protected or signing) and self.command not in ('GET', 'HEAD') else None
+        if claims is not None:
+            self.principal = 'admin' if claims.get('admin') is True else claims.get('sub')
+        if (protected or signing) and not state['key']:
            body, kind, code = b'Not found\n', 'text/plain; charset=utf-8', 404
-        elif protected and not self.authorized(state['jwt_secret']):
+        elif signing and self.command != 'POST':
+            body, kind, code = b'Method not allowed\n', 'text/plain; charset=utf-8', 405
+        elif signing:
+            wait = auth_wait(self.client()) if path != '/auth/logout' else 0
+            if wait:
+                code, document = 429, {'error': f'too many sign-in calls; retry in {wait} seconds'}
+                extra.append(('Retry-After', str(wait)))
+            else:
+                code, document = auth(path, payload)
+            body, kind = json.dumps(document, indent=2).encode() + b'\n', 'application/json'
+        elif protected and claims is None:
            body, kind, code = b'Unauthorized\n', 'text/plain; charset=utf-8', 401
            extra.append(('WWW-Authenticate', 'Bearer'))
+        elif protected and (refused := self.refusal(claims, path, payload)):
+            code, kind = refused[0], 'application/json'
+            body = json.dumps({'error': refused[1]}, indent=2).encode() + b'\n'
        elif path == '/config' and self.command in ('GET', 'HEAD'):
            body, kind, code = json.dumps(details(state), indent=2).encode() + b'\n', 'application/json', 200
        elif protected and path != '/config':
-            code, document = (users if directory == 'users' else groups)(state, self.command, path, self.request_body())
+            code, document = (users if directory == 'users' else groups)(state, self.command, path, payload)
            body, kind = json.dumps(document, indent=2).encode() + b'\n', 'application/json'
        elif self.command not in ('GET', 'HEAD'):
            body, kind, code = b'Method not allowed\n', 'text/plain; charset=utf-8', 405
@@ -578,66 +688,47 @@ class Handler(BaseHTTPRequestHandler):
        self.answer(True)

    def log_message(self, format, *args):
-        sys.stderr.write('%s - - [%s] %s\n' % (self.client_address[0], self.log_date_time_string(), format % args))
-
-
-def wanted():
-    """Where the service should listen now, as (address, port), or the reason it should not listen."""
-    configuration = read_json(CONFIG_FILE)
-    web = section(configuration, 'web')
-    if web.get('enabled', True) is False:
-        return None, 'the web service is disabled in the configuration'
-    port = web.get('port', DEFAULT_PORT)
-    if isinstance(port, bool) or not isinstance(port, int) or not 1 <= port <= 65535:
-        port = DEFAULT_PORT
-    general, interfaces, local = (section(configuration, name) for name in ('general', 'interfaces', 'local'))
-    address = vpnrouter_net.internal_address(str(general.get('mode') or 'manual'), str(interfaces.get('internal') or ''),
-                                             str(local.get('int_addr') or ''))
-    if not address:
-        return None, 'the internal address is not known yet'
-    return (address, port), ''
+        principal = getattr(self, 'principal', None) or '-'
+        sys.stderr.write('%s - %s [%s] %s\n' % (self.client_address[0], principal, self.log_date_time_string(), format % args))


 def log(message):
    print(f'vpn-router-web: {message}', file=sys.stderr, flush=True)


-def discover():
-    """Work out where to listen every DISCOVERY_SECONDS, and run the server thread there.
-
-    When the address or the port changes, the server thread is stopped and a new one listens at
-    the new place; while the service is disabled, the address is not known or it cannot be bound,
-    no server runs and the next round tries again.
-    """
-    server, thread, listening, reported = None, None, None, None
+def initialize():
+    """Load the PostgreSQL driver, make the signing key and the refresh tokens' database, retrying every INIT_RETRY_SECONDS until all are ready."""
+    reported = None
    while True:
-        target, reason = wanted()
-        if target != listening and server is not None:
-            server.shutdown()
-            server.server_close()
-            thread.join()
-            server, listening = None, None
-        if target is not None and server is None:
-            try:
-                server = ThreadingHTTPServer(target, Handler)
-            except OSError as exc:
-                reason = f'cannot listen on {target[0]}:{target[1]}: {exc}'
-            else:
-                thread = threading.Thread(target=server.serve_forever, name='server', daemon=True)
-                thread.start()
-                listening, reason = target, f'serving on http://{target[0]}:{target[1]}'
-        if reason and reason != reported:
-            log(reason)
-            reported = reason
-        time.sleep(DISCOVERY_SECONDS)
+        phase = 'loading the PostgreSQL driver'
+        try:
+            vpnrouter_auth.driver()
+            phase = 'reading the configuration'
+            cfg = vpnrouter.load_config()
+            phase = 'making the signing key'
+            vpnrouter_auth.ensure_key()
+            phase = 'creating the token database'
+            vpnrouter_auth.provision(cfg)
+            phase = 'connecting to the token database'
+            with vpnrouter_auth.database(cfg):
+                pass
+        except (OSError, SystemExit, vpnrouter_auth.DatabaseError) as exc:
+            error = f'{phase} failed: {exc.code if isinstance(exc, SystemExit) else exc}'
+            if error != reported:
+                log(f'{error}; retrying every {INIT_RETRY_SECONDS} seconds')
+                reported = error
+            time.sleep(INIT_RETRY_SECONDS)
+            continue
+        log('ready')
+        return


 def main():
    signal.signal(signal.SIGTERM, lambda *_: sys.exit(0))
-    discovery = threading.Thread(target=discover, name='discovery', daemon=True)
-    discovery.start()
+    initialize()
    try:
-        discovery.join()
+        vpnrouter_serve.serve(lambda: vpnrouter_net.web_listen(read_json(CONFIG_FILE), DEFAULT_PORT), Handler, log,
+                              DISCOVERY_SECONDS)
    except KeyboardInterrupt:
        pass
    return 0
@@ -0,0 +1,79 @@
+#!/usr/bin/python3
+"""Make a token with the admin claim for the router's web service, with vpnrouter_auth.
+
+The token is signed with the router's key and is valid for the lifetime given, without a
+maximum. It may call every protected path, and it is the only token that may change the
+admins group. The answer is the token on standard output, and a failure is a message on
+standard error. With --json the answer is {"token", "expires"} on standard output, and a
+failure is {"error": ...} there too. The exit status is 2 for an invalid request and 1 for
+any other failure. -h or --help shows the options.
+
+  vpn-router-token [--json] (--minutes N | --hours N | --days N)
+"""
+
+import argparse
+import json
+import sys
+
+sys.path.insert(0, '/usr/lib/vpn-router')
+
+import vpnrouter_auth
+
+#: The seconds of each unit of the lifetime.
+UNITS = {'minutes': 60, 'hours': 3600, 'days': 86400}
+
+
+class Failure(Exception):
+    """A request that cannot be done; status is 2 invalid, 1 other."""
+
+    def __init__(self, message, status=1):
+        super().__init__(message)
+        self.status = status
+
+
+class Parser(argparse.ArgumentParser):
+    """An argument parser whose errors are Failures with status 2, so that --json answers them as JSON."""
+
+    def error(self, message):
+        raise Failure(f'{message}; see {self.prog} --help', 2)
+
+
+def positive(value):
+    try:
+        number = int(value)
+    except ValueError:
+        number = 0
+    if number < 1:
+        raise argparse.ArgumentTypeError(f'{value!r} is not a positive whole number')
+    return number
+
+
+def main(argv):
+    # A request that cannot be parsed is answered in JSON when --json is among the arguments.
+    as_json = '--json' in argv
+    try:
+        parser = Parser(prog='vpn-router-token', description='Make a token with the admin claim for the web service.')
+        parser.add_argument('--json', action='store_true', help='answer in JSON')
+        lifetime = parser.add_mutually_exclusive_group(required=True)
+        for unit in UNITS:
+            lifetime.add_argument(f'--{unit}', type=positive, metavar='N', help=f'valid for N {unit}')
+        options = parser.parse_args(argv)
+        as_json = options.json
+        seconds = next(getattr(options, unit) * factor for unit, factor in UNITS.items() if getattr(options, unit))
+        answer = vpnrouter_auth.admin_token(seconds)
+    except Failure as failure:
+        message, status = str(failure), failure.status
+    except OSError as exc:
+        message, status = str(exc), 1
+    else:
+        print(json.dumps(answer) if as_json else answer['token'])
+        return 0
+    if as_json:
+        print(json.dumps({'error': message}))
+    else:
+        print(f'vpn-router-token: {message}', file=sys.stderr)
+    return status
+
+
+if __name__ == '__main__':
+    sys.exit(main(sys.argv[1:]))
@@ -1,7 +1,8 @@
 #!/usr/bin/python3
 """Manage the users of the router's directory from the command line, with vpnrouter_users.

-Passwords are read from the first line of standard input, never from the command line. The
+Passwords and TOTP secrets are read from the first line of standard input, never from the
+command line. The
 answer is text for people on standard output, and a failure is a message on standard error.
 With --json the answer is JSON on standard output, and a failure is {"error": ...} there too.
 The exit status says what failed: 2 the request is invalid, 3 the user does not exist, 4 the
@@ -13,8 +14,12 @@ user exists, 1 anything else. -h or --help shows the commands and their options.
  vpn-router-user [--json] passwd <name>       (password on stdin)
  vpn-router-user [--json] disable <name>
  vpn-router-user [--json] enable <name>
+  vpn-router-user [--json] unlock <name>
  vpn-router-user [--json] remove <name>
  vpn-router-user [--json] profile <name>
+  vpn-router-user [--json] otp-new <name>
+  vpn-router-user [--json] otp-enrol <name> <code>   (secret on stdin)
+  vpn-router-user [--json] otp-remove <name>
 """

 import argparse
@@ -25,6 +30,7 @@ sys.path.insert(0, '/usr/lib/vpn-router')

 import ldap

+import vpnrouter_auth
 import vpnrouter_users
 from vpnrouter_users import Failure

@@ -36,8 +42,12 @@ HELP = {
    'passwd': 'set the password, read from standard input',
    'disable': 'lock the user out of OpenVPN',
    'enable': 'let the user in again',
+    'unlock': 'end a lockout after failed logins',
    'remove': 'revoke the certificate and remove the user',
    'profile': 'print the OpenVPN client profile',
+    'otp-new': 'make a TOTP secret and its otpauth URI; nothing is stored',
+    'otp-enrol': 'store the TOTP secret read from standard input, when the code is a code of it',
+    'otp-remove': 'remove the TOTP token',
 }

 #: The help of the person options.
@@ -63,16 +73,18 @@ def parse(argv):
            for key, _ in vpnrouter_users.PERSON:
                sub.add_argument(f'--{key.replace("_", "-")}', dest=key, metavar='TEXT', required=command == 'add',
                                 help=PERSON_HELP[key])
+        if command == 'otp-enrol':
+            sub.add_argument('code', help='a code of the secret from the authenticator')
    return parser.parse_args(argv)


 def password():
-    """The password from the first line of standard input."""
+    """The password or the TOTP secret from the first line of standard input."""
    return sys.stdin.readline().rstrip('\n')


 def run(connection, options, secret):
-    """The answer of the command the options name; secret is the password of add and passwd."""
+    """The answer of the command the options name; secret is the password of add and passwd, the TOTP secret of otp-enrol."""
    command = options.command
    if command == 'list':
        return vpnrouter_users.list_users(connection)
@@ -83,21 +95,27 @@ def run(connection, options, secret):
        return vpnrouter_users.update(connection, options.name, values)
    if command == 'passwd':
        return vpnrouter_users.passwd(connection, options.name, secret)
-    return getattr(vpnrouter_users, command)(connection, options.name)
+    if command == 'otp-enrol':
+        return vpnrouter_users.otp_enrol(connection, options.name, secret, options.code)
+    return getattr(vpnrouter_users, command.replace('-', '_'))(connection, options.name)


 def text_of(answer):
    """The answer as text for people."""
    if 'profile' in answer:
        return answer['profile'].rstrip('\n')
+    if 'secret' in answer:
+        return f'secret  {answer["secret"]}\nuri     {answer["uri"]}'
    if answer.get('removed'):
        return f'user {answer["name"]} removed'
    users = answer['users'] if 'users' in answer else [answer]
    if not users:
        return 'no users'
-    head = ('USER', 'STATE', 'CERTIFICATE', 'FIRST NAME', 'LAST NAME', 'E-MAIL')
-    rows = [(user['name'], 'enabled' if user['enabled'] else 'disabled', 'issued' if user['certificate'] else 'missing',
-             user['first_name'] or '-', user['last_name'] or '-', user['email'] or '-') for user in users]
+    head = ('USER', 'STATE', 'CERTIFICATE', 'OTP', 'FIRST NAME', 'LAST NAME', 'E-MAIL')
+    rows = [(user['name'], 'disabled' if not user['enabled'] else 'locked' if user['locked'] else 'enabled',
+             'issued' if user['certificate'] else 'missing',
+             'enrolled' if user['otp'] else '-', user['first_name'] or '-', user['last_name'] or '-', user['email'] or '-')
+            for user in users]
    widths = [max(len(row[column]) for row in (head, *rows)) for column in range(len(head))]
    return '\n'.join('  '.join(cell.ljust(width) for cell, width in zip(row, widths)).rstrip() for row in (head, *rows))

@@ -110,13 +128,15 @@ def main(argv):
        as_json = options.json
        if options.command != 'list':
            vpnrouter_users.check_name(options.name)
-        secret = password() if options.command in ('add', 'passwd') else None
+        secret = password() if options.command in ('add', 'passwd', 'otp-enrol') else None
        with vpnrouter_users.session() as connection:
            answer = run(connection, options, secret)
    except Failure as failure:
        message, status = str(failure), failure.status
    except (ldap.LDAPError, OSError) as exc:
        message, status = f'the user directory failed: {exc}', 1
+    except vpnrouter_auth.DatabaseError as exc:
+        message, status = str(exc), 1
    else:
        print(json.dumps(answer) if as_json else text_of(answer))
        return 0
@@ -46,6 +46,7 @@
  "web": {
    "enabled": true,
    "port": 7081,
-    "jwt_secret": ""
+    "database": "",
+    "trusted_proxies": []
  }
 }
Author	SHA1	Message	Date
slawek	732cc8cb54	fix: the web service waits for the PostgreSQL driver instead of failing at start. Release / release (push) Successful in 25s Details	2026-10-24 10:43:51 +02:00
slawek	c0cd323a0c	feature: bootstrap diagnostics on port 17081 during the first setup. Release / release (push) Successful in 13s Details	2026-10-24 10:07:22 +02:00
slawek	2e2bacee7d	feature: the web service creates its key and database itself and shows its starting phase. Release / release (push) Successful in 29s Details	2026-10-24 09:07:28 +02:00
slawek	4f1aa282f6	fix: a failing user directory or token database no longer leaves the tunnels down. Release / release (push) Failing after 15m34s Details	2026-10-24 08:55:25 +02:00
slawek	caf0e334f0	feature: users sign in to the web service, and only members of admins manage the router. Release / release (push) Failing after 16m57s Details	2026-10-24 04:19:01 +02:00
slawek	791fc7fddd	feature: users can be enrolled for TOTP one-time passwords. Release / release (push) Failing after 15m23s Details	2026-10-24 03:00:35 +02:00
slawek	f135d776f8	fix: package upgrades no longer ask the configuration questions. Release / release (push) Successful in 18s Details	2026-10-24 02:16:19 +02:00