Jarl works out which views are reachable from the internet (match-clients
ending in a catch-all, not private-looking addresses: an internal view of a
global IPv6 prefix is still yours; a server without views: open allow-query
and a non-private listen address), with a per-view override (schema 9), and
checks public ones: rate limiting, open resolver, minimal-any, cache
snooping, open NOTIFY and transfers, version and host name, management
interfaces. Fixes are staged like any change, including rate limiting from
ISC's example in log-only mode with the secondaries exempt, then enforcing.
named logs what rate limiting limits (or would, in log-only mode, where its
counters don't move) as query-errors; Jarl routes those lines to its log
and counts them (rrl breakdown, ratelimited top lists, Prometheus
jarl_ratelimit_logged_total), shown on the new Exposure page.
A change that newly opens a public view up must be confirmed (accept_risk,
--accept-risk, a checkbox on the Changes page; 409 confirm_risk
otherwise). Findings can be acknowledged with a reason and stay visible as
info. GET /exposure, POST /exposure/fix|ack|unack, jarl exposure,
docs/exposure.md.
Also fixes query statistics skipping TSIG-signed queries ("IP#port/key
NAME" in named's log).
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
jarl changeset show/diff/apply/discard take the changeset from -C or
JARL_CHANGESET when no ID is given, and remove takes [ID] SEQ. A changeset
from JARL_CHANGESET is checked before anything uses it (staging included):
it must exist, be open and be the caller's own, so a stale export can't
stage into an applied changeset or into another user's draft that later
got its ID. Commands say on stderr which changeset they use, and apply and
discard remind to unset JARL_CHANGESET. -C and ID arguments are used as
given.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Only a release build (the commit tagged v<VERSION>, unchanged) gives a
version number. Others print 'dev <commit> (3 commits after v0.2.0-rc.1)',
from git describe (the Makefile's DESCRIBE), in jarld version,
jarl --version, jarl status, jarl whoami and the startup log; a build
without git says '(source build)'. The CLI's User-Agent and the footer's
tooltip follow suit, and /status's build gains since and ahead.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
- Daily database backups: jarl-backup.timer runs 'jarld backup', which
writes jarl-<UTC time>.db into backup.dir, checks it before naming it,
and prunes by the time in the name while always keeping the newest
keep_at_least. New backup: section in jarl.yaml; the dashboard and
'jarl status' show the newest backup and warn when it is stale or
failed. The Debian package enables the timer.
- 'jarld restore [FILE | latest]': checks the backup, shows it next to
the current database, keeps the current one aside and records the
restore. 'jarld serve' holds a database lock, so a restore can't run
under a live jarld and two jarlds can't share a database.
- Token expiry warnings from 14 days ahead: dashboard, token lists, a
daily log line, and a CLI warning via the Jarl-Token-Expires header;
'jarl token list --expiring'.
- 'jarl zone serials' and the zone page's Name servers tab: compare the
serial on a primary zone's secondaries (NS records, also-notify) or a
secondary zone's primaries with this server's.
- 'jarl zone delegation': for zones on the internet, check the parent's
delegation and glue against the zone; an undelegated zone gets a note.
- Server: Jarl/<version> header; CLI exit code 2 for failed checks.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
jarld also listens on /run/jarl/jarl.sock (socket in jarl.yaml). Requests
on it from root or jarld's own user need no token and act as an
administrator; the kernel's SO_PEERCRED says who is connecting, and
anyone else must send a token. The jarl CLI uses the socket when no server
is configured, or with -s local, so 'sudo jarl status' works on the server
with nothing set up and a lost token or password never locks root out.
Changes are audited under the system user's name via "socket".
The systemd unit creates /run/jarl (0750). Socket paths over 107 bytes
are refused with a clear message. The docs explain the socket and what
happens when a token is revoked.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
- CLI: name several servers and groups in the config file; select them
with -s/--server NAME|GROUP|all|URL or JARL_SERVER. A command aimed at
several servers is dry-run on all of them first and only runs if every
one accepts it. Named servers always use their own token.
- Reverse zones: /32 and /128 get single-address zones; networks between
octet or nibble boundaries (e.g. a /22) are covered by several zones
created together, from the CLI and the web UI.
- Fix: under jarl.service's RestrictSUIDSGID, creating a view directory
failed on chmod with the setgid bit; fall back to the permission bits.
- Web UI: the review/live switch is a small header button, and the
pending-changes bar only shows when there are pending changes.
- TSIG key wording; the keys page points to restricted API tokens.
- Docs: several servers (cli.md), a secondary server (operations.md),
tokens vs TSIG keys (permissions.md); CHANGELOG.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Jarl keeps a BIND 9.20 server's whole configuration (views, ACLs, TSIG
keys, zones of every type, records, named.conf options) in SQLite and
renders named.conf, keys.conf and zone files from it.
Every change goes through one changeset pipeline: whole-configuration
validation, named-checkconf -z and named-checkzone in a staging
directory, atomic install, rndc reconfig/reload, serial verification
with rndc zonestatus, and rollback if named does not load the change.
Changes apply one at a time by default or are bundled and applied
together. Applied changesets can be reverted.
- jarld: the daemon with the REST API and an embedded web UI, plus import,
adopt, check, resync, backup and local user recovery commands
- jarl: the CLI, including ACME DNS-01 hooks for certbot and acme.sh
- users with argon2id passwords and CSRF-protected sessions; hashed API
tokens with granular grants and roles; an audit log
- systemd unit, sysusers.d/tmpfiles.d files, Makefile
- unit tests, integration tests against a real named, scripts/e2e.sh
- docs/ guides, README and CHANGELOG
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
adamandClaude Opus 5.5
abrown38andClaude Opus 5.5