Answers, result codes, record types, zones, blocking, rate limiting and the top lists are also counted per view, from named's response log. The Statistics page has a toggle per view with All and None, remembered per browser; jarl stats --view and ?views= on the API. named's own counters stay for the whole server, and the page says so.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
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>
New metrics: section (enabled, listen, allow, token). /metrics is served on
the main listeners, the local socket (root and jarld's user) and optional
metrics-only addresses; a scrape needs an allowed address or the bearer
token. Exports health (named up, problems, drift, backups, blocklists,
DNSSEC phases, token expiry) and, with stats on, named's and Jarl's query
counters since jarld started plus named's process and cache gauges.
docs/metrics.md has example queries and alerts checked with promtool.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Added:
- Server statistics: named's memory and CPU (from /proc), cache size,
sockets and client counts against their limits (statistics channel,
rndc status), host memory and load, and free disk space. Memory and
CPU are kept over time as gauges (sum and sample count, averaged in
timelines). Statistics page Server section, jarl stats server,
GET /stats/server.
- jarl record add|set|delete|disable|enable --view a,b | all: the same
change in each view, applied as one changeset.
- Blocking can be paused until resumed (jarl block pause indefinitely,
"Until I resume", {"indefinitely": true}).
Changed:
- List options (allow-transfer, forwarders, ...) get their braces added
when given without them.
- The dashboard lists named-checkconf's warnings about the running
configuration; the CLI prints them after an apply.
- Stats page: tile and card rows fill their width, and the queries and
failed answers charts share margins so their time axes line up.
- Dashboard: options have their own tile.
- 24-hour times in the web UI.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
After a reconfig that changed named's views, the cache and resolver
counters (summed over views) went down, which was taken for a restart:
every counter then counted its total since named started for one minute.
Only a new boot time means a restart now; a counter that goes down
counts 0.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
jarl block pause|resume (and the Blocking page) lets every name through
for a while; jarld resumes on time. The lists stay loaded: the rules zone
just passes everything. Statistics gain the clients with the most blocked
queries and blocked queries on the queries chart. The view choice in the
blocking forms is a dropdown.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
jarld downloads blocklists (Adblock, hosts or plain domains), checks them
daily and reloads named only when a list changed; your own block and
allow rules take precedence, and the most specific rule wins. Lists and
rules apply per view (default: every view that offers recursion). Web UI
Blocking page with name checks, jarl blocklist and jarl block, the
/blocking API, blocked counts in the statistics. Database schema 6.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
The zones list shows a padlock after signed zones, linking to their
DNSSEC tab. Client lists in the statistics show the host name Jarl's
records have for each address (PTR, else A/AAAA); no DNS queries.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
make deb now builds two packages: jarl-cli (the CLI, its completion and
the ACME hooks, no scripts) and jarl (the server), which depends on
jarl-cli of the same version.
Pushing a v* tag runs .gitea/workflows/release.yml: it tests, builds the
tarballs, the CLI for Linux, macOS and Windows (make cli-dist) and the
.debs for amd64 and arm64, then scripts/gitea-release.sh creates the
Gitea release and uploads the .debs to the Debian package registry, so
servers install and upgrade with apt. The stats package's inode lookup
is split by OS so the CLI builds for Windows.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Off by default (stats in jarl.yaml). When enabled, Jarl adds a loopback
statistics channel and a response log to named's configuration (merged
into existing logging/statistics-channels statements) and switches
response logging with rndc responselog, because the responselog option
only counts when named starts. jarld polls named's counters every minute
and tails the response log (resuming after restarts and across
rotation). It stores aggregates, never individual queries, in a separate
SQLite database: per-minute, hourly and daily counters (queries, result
codes, record types, views, zones, cache hits, upstream latency, DNSSEC)
and hourly/daily top lists (names, clients, NXDOMAIN and SERVFAIL names,
refused clients), each with its own retention. Clients can be anonymized
or excluded.
- Web UI: Statistics page with KPI tiles, queries and failed answers over
time, a weekday x hour heatmap, top lists and breakdowns; SVG charts
with hover/keyboard readouts, a table view and light/dark palettes
validated for colour-vision deficiency.
- CLI: jarl stats status|summary|top|timeline|heatmap|breakdown.
- API: /api/v1/stats/*, behind a new "stats" permission resource.
- jarld brings named's configuration up to date with jarl.yaml at
startup (engine.SyncConfig), never overwriting hand-edited files.
- resync now refuses to overwrite drifted files without force, as
documented.
- tmpfiles.d creates /var/log/named (allowed by named's AppArmor profile
on Debian 13 and Ubuntu).
- The e2e harness runs named with -f like the Debian unit.
- docs/stats.md, plus API, CLI, configuration, permissions and README
updates.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
adamandClaude Opus 5.5
abrown38andClaude Opus 5.5