A Cache section between the pools and the schedules. One card for the ARC:
size against its ceiling with the target and floor, sparklines of size and
hit rate from the sampler's history, what it holds (MRU/MFU/dirty, data vs
metadata, compression, the ghost lists), the prefetcher and dbuf cache, the
memory-pressure counters, and the since-boot breakdown by demand/prefetch
and data/metadata. Then a card per pool for L2ARC, SLOG, special vdev,
dedup table and spares — the devices and their fill where they exist, and
where they don't, plainly none and what that means (the ZIL writing to the
pool's own vdevs, with how much it has written).
Live numbers come from the sampler, which reads arcstats every tick and
publishes zfs.arc.* sensors; the fuller picture is read on request, with
zpool list -Hpv and zfs get run without sudo and cached for thirty seconds.
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
Every DRM device with a real GPU driver gets a box under the cpu box, laid out
like it: the busy curve, a meter for the device and one per engine, temperature,
clock, power and fan where the card has them, and the processes holding it open
with their share. The process table gains a Gpu% column and sort.
Device busy is Intel's RC6 idle residency inverted (what intel_gpu_top shows,
and readable without root) or amdgpu's gpu_busy_percent. Per-process shares are
DRM fdinfo deltas, as nvtop reads them; other users' fdinfo needs root, so a
daemon running as a user keeps one root child (hamlab drm-clients) over a pipe
rather than paying a sudo a second. Everything is also a gpu.* sensor.
What cannot be measured is left out, not shown as zero: an Arc A380 on i915 6.12
has no temperature sensor, an integrated GPU wears the package temperature
marked *, and Asahi has neither a busy counter nor fdinfo yet.
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
Per machine like every other tab: the interfaces as the kernel has them
(addresses, MAC, MTU, negotiated link speed, carrier, totals, live rates from
the sampler's own sensors), the default route and resolvers, what the internet
sees of this connection, and a speed test with the last fifty results kept.
The test runs in the daemon, not the browser, so the figure is the line of the
machine in the switcher — the router's, if you switch to gate — and it carries
on when the phone locks. It measures against Cloudflare's public speed
endpoints, so there is nothing to install on any host, which is what makes it
work the same on Debian and on OpenWrt where flash is scarce.
What the measurement does and does not count:
- Several streams, HTTP/2 deliberately off. One stream over one connection
measures a congestion window, and HTTP/2 would put every stream back onto
one connection and measure the same thing.
- The first second of each phase is thrown away as the ramp.
- A phase ends on 64 MiB as well as on four seconds, so a fast line says what
it is without moving a gigabyte. A minimum window keeps a very fast line
from filling the cap before there is anything to divide by.
- Latency is the round trip less the far end's own Server-Timing. A zero-byte
response is mostly Cloudflare's worker thinking: left in, it reported 31 ms
for a line the edge itself measures at 3.5. Jitter is the mean change
between consecutive probes, so one slow probe cannot run away with it.
- The far end rate-limits by volume. So: a 20 s cooldown counted from the last
attempt rather than the last success, streams that retry twice, the rate
computed over the window in which bytes were actually still arriving, and a
plain sentence when it was refused. Coasting to the end of the clock with
nothing arriving reported a third of the truth.
The last result is published as net.internet.* sensors, each stamped with when
the test ran, so the line can be a widget, a graph, an alert or a Prometheus
series like anything else. The public address is refreshed on its own quarter
hour, which is worth having here: it is a DHCP lease and it moves.
Tests are offline — the window, the byte cap, the refusal and the cancellation
run against a local server; the route parser's byte order is pinned by a case
that would have caught the bug it did. The live measurement is behind
HAMLAB_NET_TEST=1.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
An Apple SSD presents three namespaces and the kernel lists each as a whole disk, so the
dashboard was showing one drive as three. The Disks page has always folded them; now the
dashboard, the Top tab and the disk.* sensors do too.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
The process table's columns fit the box they are in, the rate control saves, curves are
timed on the viewer's clock and keep to the stride, the power icon is drawn, and a young
history draws in from the right instead of being stretched.
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
Boxes with their name in the border, dot curves that climb green to red, per-core meters and
temperatures, memory and filesystems, per-interface traffic and the process table. Drawn from
the sensor stream and /api/procs, so it works on every machine the console can see.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Each machine keeps 300 points per curve whatever the window, so memory doesn't follow it;
a wider window means wider points, each the average of its bucket rather than one instant
out of it. The hub passes the setting to every machine it manages.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
- service.<unit>.active|state for services that are enabled, up or failed
- containers.* and container.<name>.running|state|health|mem|cpu (cgroup reads, no docker stats)
- sensors:read tokens can be limited to ID patterns, enforced on every endpoint and,
through the hub's filter header, on managed machines; the header can only narrow
- /api/fleet/sensors/query: one shaped query, every machine, one request
- the sampler waits what's left of the interval, so 250 ms means 4 a second
- Monitors asks for the families it draws instead of *
- unpowered hwmon sentinels (-93 °C) are not temperatures
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
- stock Monitors modules fed by one sensor stream, the catalog and one batch of
histories (GET /api/sensors/history?ids=…), through sampleAdapter.ts
- the stream closes while the page is hidden and refills its curves on return
- Disks tab live rates from the stream
- new sensors: hamlab.sample_interval, ups.last_event
- stream: asking for about the sampling rate means every sample
- NUT runtime rounded at the source
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
- sensors:read tokens open the sensor endpoints (here and on managed machines
through the hub) and nothing else; full tokens are never accepted from a URL
- /api/sensors/stream: a shaped query that keeps answering, over SSE
- /metrics: Prometheus text format
- services, updates, ZFS and SMART as background status sensors
- widget library over the sensor API; custom widgets on the dashboard
- /embed/widget: one live widget in its own bundle, the only frameable path
- Dashboard tab is now Monitors
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
Every number the collector samples becomes a sensor with a stable dotted ID
(cpu.core.3.usage, fan.nct6775.fan2, fs.root.used_pct). Clients discover them
from the catalog and ask for exactly what they want, including as a shaped
query: send the object you want, get it back with the values filled in.
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
