The power limit and 810 MHz memory lock were a host setting, applied at
every boot by nvidia-power-limit.service regardless of what the GPU was
doing. A host switched from mining back to LLM inference kept the lock
with the miner stopped, and inference is memory-bandwidth bound: a 27B
model on 2x RTX 5090 decoded at 3.8 tok/s on a build that had measured
21.9 tok/s. Nothing reported it — a -lmc lock is not a clocks event
reason, and utilisation still reads ~98%.
The profile now follows the workload:
- nvidia-power-limit.sh takes a mode. `miner` applies the configured
profile as before; `default` resets each card to its own default power
limit and memory clock.
- nvidia-power-limit.service runs `default` at boot. The miner applies
`miner` in ExecStartPre and `default` in ExecStopPost (quantus/miner).
- A drop-in for neuron.service makes it conflict with the miner and
reset to default before starting, so either one can take the host with
a single systemctl start and the clocks follow.
- infra-setup.sh installs the drop-in, always writes the profile config
(so removing a host clears it), and applies whichever profile matches
the running workload instead of restarting the boot unit under a live
miner.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01XViS3RVfGTRaRSopLN68MA
A per-node rpc_extra_sources list of mesh names, each resolved on the node
into a /32 firewalld rich rule for quantus-node-rpc, beside the site-subnet
rule. bob gets oolon.kosherinata.internal: the kosherinata edge proxy that
publishes the node as wss://quantus.blackbeard.observer for the blackbeard
wallet (blackbeard/wallet #53). The rule is already applied on bob by hand
with the same command; this makes the next deploy agree with it.
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_014ftBXYuba8ARhQeF74oUgW
Upstream released v1.0.0-gm and v1.0.1 on mainnet day; the fleet has run
0.11.1 from the published mainnet.json since the hand cutover. The node-side
delta from 0.11.1 is small: the mainnet spec is embedded (`--chain mainnet`),
command.rs adjusts four lines, and the miner server and consensus client are
untouched. The release notes state any node from v0.11.1 on is valid, so this
is a binary swap on the same chain id, spec file and credentials.
This is also the first push-triggered node deploy since the cutover, which
is the point: prove the CI path upgrades bob without hand intervention.
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01YJp25VomsWVZDPM9Fwmhnq
Mainnet cut over 2026-09-09 09:01 UTC. The spec (sha256 81f048c0…, id
"mainnet", genesis 0xfb5487c0…) is committed under asset/chainspec/ and
the deploy env carries the spec path, its checksum, the expected genesis
and the seven foundation bootnodes, so push-triggered deploys reproduce
the hand cutover exactly and validate asserts the genesis from now on.
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01CBgs2nSi4H2mdh8kD8vMX5
The mainnet package arrives as files (spec, bootnodes, genesis, checksum
file, readme) with the node staying on 0.11.1, so the spec is committed
under asset/chainspec/ after verification and the deploy takes a repo
path as well as a URL. CHAINSPEC_SHA256 / chainspec_sha256 re-verifies the
file in fetch. The fetch job now checks the repo out.
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01CBgs2nSi4H2mdh8kD8vMX5
Mainnet call 2026-09-09: beast mines. The node's firewalld rich rule for
the QUIC miner port and Prometheus's miner targets gain beast, matching
quantus/miner's deploy matrix.
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01CBgs2nSi4H2mdh8kD8vMX5
Mainnet launches 2026-09-10 09:09 UTC with the chain spec published as a
file before any binary embeds it. The node deploy can now run from
/etc/quantus-node/chainspec.json (CHAINSPEC_URL or the chainspec_url
input), asserts the started node's block-0 hash against EXPECTED_GENESIS
(required with a spec: it is the only defence against a look-alike), and
prints the spec id, which names <base-path>/chains/<id>/ and is what the
miner deploy must be pointed at; validate checks the miner credentials
exist under it. BOOTNODES are dialled and held as reserved peers.
The unit raises --out-peers to 32 and --in-peers to 64: bob sits at the
default cap of 40 peers today, and launch is a propagation race
(quantus/chain#1). infra-setup grants the runner the chain spec path
(applied to bob 2026-09-09).
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01CBgs2nSi4H2mdh8kD8vMX5
Three sources of author-label churn split the per-author hashrate series
into fragments (lair/quantus#11 follow-up):
- Attribution state lived in memory, so every restart stripped every
author of its name for a few blocks. Votes, held names and the peer-id
to name map are now saved to --state-file (the service uses
/var/lib/quantus-arena/attribution.json) at most every 10 s and
restored at start.
- A name dropped back to the preimage whenever confidence dipped under
0.6. A held name now stays until its node has no vote left in the
20-block window or another node out-votes it outright.
- A duplicate name got its peer-id suffix only once a second node with
that name was attributed, re-labelling the first. Any name used by more
than one node on the feed now always carries the suffix.
Dashboard: the hashrate-by-author panel no longer spans nulls, so a label
change shows as a gap rather than a slope drawn between two points.
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01CBgs2nSi4H2mdh8kD8vMX5
New interval variable 'author window' (auto: about a quarter of the
range, never under 2 h; or pinned 1h..2d) replaces the fixed 6 h trailing
window in the hashrate-by-author panel, so a week view smooths and a
short view resolves.
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01CBgs2nSi4H2mdh8kD8vMX5
After a restart the window started empty, so authors who had not sealed a
block since had no leaderboard row source, no author label series and no
timing caps until they authored again; on the dashboard that showed as
blank author cells for the smaller miners. The first poll now fills the
window (3600 blocks, 15 s against bob) from chain history, window only:
the authored counter is untouched so increase() sees no phantom burst, and
no tip timing is observed for blocks not watched arriving. Capped at 180 s
of wall clock.
The leaderboard's top-20 also drops rows with zero blocks in the range;
they were stale series with nothing to show.
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01CBgs2nSi4H2mdh8kD8vMX5
Each author's estimated hashrate over time: share of blocks authored in
the trailing 6 h times the 6 h average network hashrate, legend by the
telemetry node name where attributed, else the abbreviated preimage.
Authors under 10 MH/s are left out. Makes pool growth visible as a trend
rather than a leaderboard snapshot.
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01CBgs2nSi4H2mdh8kD8vMX5
The exporter emits quantus_author_node_info only for authors in its rolling
block window, which is rebuilt from live blocks after a restart, while the
leaderboard covers the dashboard range. Authors outside the window showed a
blank author cell. Query G now takes the current series and, for preimages
without one, the last label set seen in the range (one per preimage).
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01CBgs2nSi4H2mdh8kD8vMX5
Telemetry node ids are per connection: a node that restarts or renames
comes back under a new id, so votes keyed by node id split across the old
and new identity, and with a lifetime attempt count the new identity had to
outvote the whole history before the name came back. Votes are now keyed by
peer id (stable across restarts and renames; node id only when telemetry
gave no peer id), kept for the last 20 blocks per author, and displayed
with the latest name telemetry reported for that peer, so a rename shows
within a block and a restart within a few. Lifetime attempts/attributed
counters are unchanged.
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01CBgs2nSi4H2mdh8kD8vMX5
The exporter ingests the feed's ImportedBlock (code 6) reports and keeps,
per block hash, the first reporter and its lead over the second. Each tip
block is joined with that after 8 s; a reward preimage gets the node's
telemetry name once at least 3 of its blocks were attempted and 60% or more
were first reported by the same node. Emitted as
quantus_author_node_info{preimage,display,node_name,peer_id,self} confidence
with attempts/attributed counters per preimage; display is the name, or
the abbreviated preimage (first 8, last 4) when no confident attribution.
Measured today: leads are 40-100 ms for authors on the feed (the network is
better connected than when 200-700 ms was measured), so the lead threshold
is 20 ms and identity consistency carries the attribution. Validated for
ten minutes against bob: ours -> baba-gorchitsa, 0x2ccf39 -> quanpool-planck,
GUI miners -> QuantusMinerGUI at confidence 1.0.
Dashboard: the leaderboard joins quantus_author_node_info and shows the
display label as "author" in place of the reward preimage; the "ours" column
is gone, so the panel is fit to share. ARENA_DEBUG=1 turns on debug logging
(per-block attribution decisions).
Closes#11.
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01CBgs2nSi4H2mdh8kD8vMX5
Author timestamps are set when the node builds the proposal, right after
the previous block arrives, so gaps between author timestamps measure
proposal timing and flagged our own honest node at 2.8x the network's
quick-follow share. Withheld blocks show in arrival timing instead:
quantus_block_arrival_gap_seconds uses the exporter's observation time
(poll interval now 1 s; telemetry first-seen replaces it under #11). The
leaderboard's quick-follow column and the Author timing panel now use it,
with a 2 s window to sit above the poll resolution.
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01CBgs2nSi4H2mdh8kD8vMX5
For a Poisson block process the 10th-percentile gap of an honest author is
about one second by construction, so a fixed threshold flagged everyone.
quick-follow x is the share of an author's blocks arriving within 1 s of
the previous block divided by the network's share: 1 is normal, 2 or more
is withheld-block behaviour; needs 10 samples. Bucket boundaries are
selected as le="1.0" (Prometheus 3 normalises them to float strings).
Threshold bases are transparent so authors without samples show blank
cells, not red or blue.
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01CBgs2nSi4H2mdh8kD8vMX5
est. MH/s as a bar on its own scale; gap p10 and drift median as
threshold-coloured cells (low gap and positive drift are the suspicious
directions); a burst ratio column, consecutive self-blocks over the count
expected from the author's share, coloured amber above 2 and red above 4
and shown only where the expected count exceeds one.
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01CBgs2nSi4H2mdh8kD8vMX5
After the multi-query join Grafana ranged the gauge cell across every numeric
field in the frame, so 31 blocks was drawn against a 675 MH/s maximum.
fieldMinMax on the blocks column restores the per-column scale. The gap and
drift quantiles are filtered with expr == expr so authors without samples in
the range show a blank cell instead of NaN.
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01CBgs2nSi4H2mdh8kD8vMX5
Estimated hashrate is the author's share of blocks over the range times the
network hashrate estimate: statistical, roughly ±10% at a few hundred
blocks, and labelled as such. The timing columns join the new per-author
histograms (lair/quantus#12) on the preimage; outer join filtered to the
top-20 rows from the blocks query so authors without timing data still
appear with blanks.
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01CBgs2nSi4H2mdh8kD8vMX5
Two signals that expose what other miners are doing, both from data the
exporter already fetches plus one chain_getBlock per tip block:
- quantus_block_gap_seconds{preimage}: histogram of the time between a
block and the previous one, by the author of the later block, from the
authors' own timestamps (the Timestamp::set inherent, pallet 1 call 0,
decoded from the first extrinsic). Bursts in the lowest buckets for one
author right after others' blocks are withheld blocks.
- quantus_block_timestamp_drift_seconds{preimage}: block timestamp minus
the time the exporter first saw the block. Negative is propagation and
poll lag; positive is a future-dated block, which the Homestead-style
retarget rewards slightly (the inherent allows 30 s).
- quantus_consecutive_self_blocks_total{preimage}: blocks authored directly
after the same author's previous block.
Same top-N cap as the leaderboard, with "other" merged and "all" for the
network. Validated live against bob for three minutes. Dashboard row
"Author timing" with p10 gap, median drift and consecutive-block rate.
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01CBgs2nSi4H2mdh8kD8vMX5
grenade
