docs/PLUGIN_DEVELOPMENT.md referenced a nonexistent hbd/plugin.py and
hbd/plugins/ directory; the real module is hbd/client/plugin.py with
plugins under hbd/client/plugins/, as README.md and the actual code
already use. Anyone following the doc would hit ModuleNotFoundError.
foot.html was never included by any template (confirmed via git grep
and template audit) despite being actively maintained; wired it into
about.html. menu.html was fully commented out and unreferenced, so
removed it.
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
CMD packets arrive as unauthenticated UDP datagrams, yet every hbc client
executed the shell command they carry without any opt-in. Add an
allow_remote_command config key, default false: when off, the command is
logged and refused with "Refused: allow_remote_command is false" (visible in
the server event log under the command service), and subprocess is never
reached. When on, the client warns at startup that it will execute CMD
packets.
Applied to all four clients that handle CMD — hbc, hbc_windows.py,
hbc_mini.py, and the C hbc_mini — since gating only one leaves the others
wide open. The C client reads the same key from ~/.hbc.json and needs a
rebuild to pick it up. UPD (self-update) is unchanged.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
A (host, service) pair is flapping once it exceeds flap_count warning or
critical notifications within flap_interval minutes. The notification that
trips the state carries "Now flapping!! No more messages!" and every later
one is dropped, including RECOVER. The state ends silently flap_interval
minutes after a RECOVER, provided no further alert arrived meanwhile.
Hooked into notify.send_notification, the single choke point for channel
delivery, so only outbound notifications are suppressed — eventlog keeps
recording, leaving the journal and /log with the full history of the flap.
Threshold alerts key on their metric path, so a flapping disk check cannot
silence a CPU alert; connectivity, boot and shutdown events key on the host
itself. State lives at module level in flap.py and is never pickled.
Flapping pairs surface in Host.stateinfo() and render as an amber badge on
the live dashboard. Config: flap_count (5), flap_interval (10 minutes),
0 in either disables; both editable on the settings page.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Replaces the 11-column table with severity-sorted record rows: host,
combined status chip, latency, alert chips, time-in-state, and the most
recent alert message. Per-family address/state/latency/last-change move
to a hover card on the status chip and permanently to a new Connectivity
table in the Host Overview info section (host info API now includes
connections). Last-alert is seeded from /api/0/alerts and kept fresh
from the event stream; the event log below uses the same row idiom with
its filters intact. WS reconnect state shows in the toolbar instead of
a modal.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01NfPpSpccTWBfZg1FTveyaU
Extract the settings redesign's tokens and components into shared
static/hbd-ui.css + hbd-ui.js and dedupe settings.html against them.
About becomes yaml-key sections with kv rows; Alerts gets stat tiles,
chip filters, and alert record rows (same fetch/ack logic); Host
Overview keeps its DOM and live-update JS but is re-skinned to the
token system with a page toolbar. Live Dashboard intentionally
untouched pending its own redesign.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01NfPpSpccTWBfZg1FTveyaU
Sticky offsets (toolbar, rail, mobile chips, anchor margins) derive from
the measured nav/toolbar heights via CSS variables, since the fixed nav's
height varies with viewport width.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01NfPpSpccTWBfZg1FTveyaU
One idiom for every section: mono identifier, badges, fact chips, inline
accordion editor. Scalar config sections merge into one grouped 'server'
section. Sidebar becomes YAML-key nav (mobile: horizontal chips); sticky
toolbar carries pending-changes state and Publish. Staging, publish,
channel CRUD, and permission gating are unchanged — same payload shapes
and endpoints.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01NfPpSpccTWBfZg1FTveyaU
The users-section loop reused 'username' as its loop variable, overwriting
the requesting user's name so host/threshold filtering compared against the
last user in the config. Rename to 'uname' and add a regression test with a
user that is not last in the users dict.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01NfPpSpccTWBfZg1FTveyaU
Setting args.boot = False after sending the boot message was dead code: the
SIGHUP restart re-execs via os.execv(sys.argv[0], sys.argv) with the original
argv, so the re-launched process re-parsed -b and announced a boot again. The
exiting process also sent a spurious shutdown (send_shutdown armed by -b), so
each config reload looked like a host reboot to the server.
Strip -b/--boot from the argv passed to execv, and skip the shutdown message
in cleanup() when dorestart is set. A real SIGTERM/SIGINT shutdown still sends
shutdown as before.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
The server inferred each plugin's collection interval from the gap between
the last two received PLG samples, then expired data at interval * 3. After
an outage this guess was wrong: the request_update re-send of collect-once
InfoPlugins produced two close samples, yielding a tiny inferred interval
that purged permanent info data minutes after recovery.
Clients now declare each plugin's interval in the PLG message (_interval).
The server uses it directly (from the first post-recovery sample), expiring
at interval * 3 so live data survives at least two full intervals; interval
0 marks collect-once InfoPlugins that never go stale. Legacy clients omit
the field and fall back to the previous inferred-gap behavior.
Adds _interval to all three clients: hbc, hbc_mini.py, hbc_mini.c.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
data.msgs grew without bound (eventlog appended forever, also persisted in
the pickle), and ws.py replayed the entire history to every dashboard
client on connect - one JSON frame per message - making page reloads
progressively sluggish. Cap the buffer in eventlog() (configurable via
msg_buffer_size, default 500) and slice the WebSocket replay to the last
30, matching the window the HTTP render already uses.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Host.plugin_timers holds asyncio TimerHandle objects (and their lambda
callbacks) which are not picklable, causing the 5-minute state save to
fail with "Can't pickle local object reset_plugin_timer.<locals>.<lambda>".
Add Host.__getstate__ to reset plugin_timers to {} before pickling,
mirroring Connection.__getstate__; timers are recreated on the next PLG.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
- Use `is not None` for pathconf values so 0 is not silently dropped
- Broaden connectivity prefix check to catch bare "connectivity" key
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
- Path traversal: confine avatar file serving to avatar_dir (defaults to
config file directory); validate on both read and write
- UDP owner injection: server-configured owner now takes precedence over
UDP-supplied value, matching the documented intent
- Open redirect: reject non-relative next= values after login
- Stored XSS: enable Jinja2 autoescape on all template environments;
add escHtml() helper in live.html and apply to all innerHTML sinks
sourced from network data (host names, addrs, states, log messages)
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
These entries are set by the connection state machine, not by threshold
config, so they have no threshold entry and were being deleted on every
startup. Guard them explicitly so overdue/down alerts survive the purge.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
restore_connection_timers now calls _set_connectivity_alert("CRITICAL")
for DOWN, OVERDUE, and UNKNOWN connections, ensuring alerts are present
even if hbd was shut down before the transition callbacks recorded them.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Moves the plugin-state purge from the boot flag to the UP transition,
so stale history and alerts are cleared on any reconnect (reboot, or
recovery from overdue/unknown) not just detected reboots.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
When a PLG message arrives with fewer keys than the previous sample,
alert states for the missing metrics are removed immediately. Handles
nagios checks removed from configuration while the runner plugin continues.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
live.html used host.raw_name which stateinfo() never included — the
hash was always empty. Use host.name (the raw hostname stateinfo()
does include). Also exclude plugin_timers from stateinfo() to prevent
asyncio handles from breaking jsons().
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Avoids false-stale firing for slow plugins (e.g. nagios_runner at 300 s)
when the heartbeat interval is much shorter. On the first sample cancel
any leftover timer; arm the 3× stale timer only after the second sample.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
- hbdclass: add per-plugin stale timers; clear history and alerts after
3× heartbeat interval with no PLG data received
- udp: wire stale timer on every PLG message via _make_plugin_stale_callback
- http: persist new OAuth users to config file on first login
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Y axis now fits the actual data range with 10% padding rather than
fixed 0-100%. Grid lines use nice tick steps (1/2/5/10 × magnitude).
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Renders an SVG line chart above the CPU Usage row using all available
history samples (up to 100). Color adapts green/orange/red by load level.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Replacing head -1 (and the broken head -2|tail -1 attempt) with grep -m 1
stops grep after the first match, eliminating the SIGPIPE that caused exit 141.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Theme preference stored in localStorage (auto follows the OS setting).
The chosen data-theme attribute is applied synchronously in <head> to
avoid any flash of unstyled content. CSS custom properties handle all
surface, text, border and input colours across every page. The
Appearance section on the profile page lets each user switch modes.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Send history messages newest-first from the server, tagged with
history=True so the client appends rather than prepends them, avoiding
reverse-chronological display on initial load.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Hosts with watch: false in config no longer appear in the Alerts page
or nav bar alert counts. Events still appear in the Log of Events.
Hosts without a config entry default to watch: false.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Adds structured form fields for nsupdate_bin, rndc_key, and dyndomains
(comma-separated list). Wires list-type editable fields through the
generic stageFormSection path and adds DNS support to
apply_structured_section in configio.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
The key had drifted below [project.urls], making setuptools interpret it
as a URL entry and failing validation.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
The default zfs_monitor.*.status threshold used operator '>' with warning=1,
so a DEGRADED pool (status=1) never alerted (1 > 1 is false) and a FAULTED
pool (status=2) only triggered WARNING instead of CRITICAL.
Fix the operator to '>=' in THRESHOLD_DEFAULTS and the example config.
Also adds a per-metric grace period override (ThresholdConfig.grace) so
individual thresholds can bypass or shorten the global grace delay. Alerts
with grace=0 fire immediately on state change rather than waiting for a
second collection cycle. Sets grace=0 on zfs_monitor.*.status so pool
degradation alerts fire on the first data report after the event.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
- Remove rndc-key from tracking, add to .gitignore
- Move async_sms_send.py, demo_threshold.py, nagios_bad.sh to scripts/
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
- tests/test_threshold.py: has proper pytest test functions
- scripts/test_*.py: manual run scripts with no test functions
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Replace the missing requirements-dev.txt reference with extras = dev,
which installs the [dev] optional dependencies declared in pyproject.toml.
Also remove skipsdist so tox installs the package before running tests.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Replace the previous README with documentation derived from reading
the actual code, including a new section covering the C client
(scripts/c/hbc_mini.c).
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
HTTP config-mutating endpoints (publish, rollback, channel CRUD, user
self-update) were calling config.reload() directly, which only refreshed
the in-memory config dict. This skipped re-applying host.dyn/host.watched
flags to live Host objects, so enabling dyndns via the UI had no effect
until a SIGHUP was sent.
Wire a reload_callback through http.start() that calls the same
reload_configuration() function used by the SIGHUP handler, ensuring
host attributes, notify module, users, and threshold checker are all
updated on every config publish.
Also fix unmatched quote in udp.py f-string log message.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
