A round added a row a group whenever 100k articles or 15 minutes had passed,
tuned for a tail leased within minutes; with nothing leasing the tail the rows
piled up (127k planned tail ranges of a median 14k articles, 97% adjacent). Now
a round grows the group's last range when it is a planned, unspooled, untaken
row at the tail's priority, up to RANGE_ARTICLES, and only then adds rows; a
range taken meanwhile gets the slice as a row. ranges:merge-tail (--dry-run)
packs adjacent planned, uncaptured, unspooled, untaken tail ranges of a group
and priority into ranges of RANGE_ARTICLES under the lease lock.
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
A plain stop and a chunk that doesn't inflate already posted the batch so far,
so the stop snapshot stood at a stored resume point; a failed stream (the
stalled one's "Peer disconnected") skipped the post, left the batch open and
lost the window with the snapshot refused. Now only a batch the coordinator
refused stays open.
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
The corpus fill claims chunks of a leased range ahead of the leased run (the
prefetcher's claim keys are per range and chunk), stores them in the corpus
only, and the run leaves them there. The stream looked at the corpus once per
chunk and then polled the spool only: when the store came seconds after that
look, the stream waited its whole hour and the worker lost the group and its
open batch (three groups on 2026-10-27, LOG.md (118)). Now the wait loop checks
both.
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
The lease bounds alone don't select a window: the history planner's oldest
buckets are priority 0-2 as well, so the window worker at 0..0 also took 2026
history at 0 (25k ranges in 40 groups, fetched live) while 1,739 ranges of the
2016 corpus window sat at voyager 2's leftover priorities (done when the window
was marked, planned again by the reset). With LEASE_CAPTURE_ONLY the lease, next
and the prefetcher's run-ahead take capture ranges only; the run-ahead now also
keeps to the lease bounds. Test: a group with history at 0 only isn't leased,
the history range after the window isn't handed out.
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
A live reply cut short (the server's compressed data ending at ".\r\n" at a
read boundary; the spool keeps what Corpus.write refuses) made the worker give
the group back without a snapshot, losing the window and the batch since its
last post (six times in the first 40 minutes on homelab). Now the batch before
the chunk goes out, the chunk is reported (fetched again, plain the second
time) and the range is asked for again from that chunk (RangeOutcome::Again):
the stream waits for the new copy and the window stays. Test: the coordinator
hands the same range back at its resume point after a corrupt report.
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
A worker thread leases a group and takes its ranges in article order through
one carried window; the coordinator keeps one cursor a group (group_leases,
migration 6500) and hands out the next range, the rules and nothing about
where the chunks are: the range stream serves a stored range's chunks from the
corpus file and a live one's from the spool. The window's snapshot (bincode:
the frontier, the open files, releases and albums, the upload numbers, the
batch-counter table) goes to the coordinator every EDGE_SNAPSHOT_MINUTES and
when the group is given back, and the next holder restores it; a takeover
after LEASE_TTL_SECONDS plans the ranges from the snapshot's article again.
The prefetcher runs ahead of held groups' cursors and skips chunks the corpus
has. LEASE_MIN_PRIORITY..LEASE_MAX_PRIORITY bound what is leased. edge-node
holds groups by default; --ranges keeps the old per-range path.
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
deploy.sh and the nightly resume leave the edge stopped while .cache/no-edge
exists (range by range it would split releases at every edge with no merger
left). deploy/replay-window.sh takes DB and OUT and passes
REPLAY_INTO_CATALOG=1, which lets replay:corpus write a database not named
*_replay or *_scratch (never the test one).
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
pending_ladder.ts claims the oldest waiting pending releases' next target
(par2, archive, lone, 7z-tail in that order) as at least PENDING_SHARE of each
claim; a usable name (the largest named file of a par2 set or a listing, or a
lone file's yEnc name, neither packaging nor random) hands the release to the
namer, no name the next target or the end of the climb; climbs count on
pending_signatures per (tool, group, poster) and (tool, group), and a signature
with 100 climbs and 0.5% named or fewer is dropped: its waiting releases are
discarded without a probe, the rules' sample excepted. pending_namer.ts reads
a named release's articles again from the corpus or by XOVER and writes it to
the catalog under the pending key (source weak-stem, the name as inner name)
with its chunk and candidates. Migration 6400; config PENDING_*.
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
An unnamed release (no readable file name) is written to pending_releases
(migration 6300, the HDD) with its keys, counts, article span, upload tool
and probe targets, instead of releases, release_texts, posters and a chunk.
The edge sends it with an empty payload and its tool; the batch store splits
a batch by readable. The pending ladder (probes cheapest first, signature
verdicts, the move into the catalog) comes next. The replay report shows the
pending releases.
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
A release the window closed on with one file named by its stem or post, the
name with a number and no archive, volume, sidecar or random name, waits in
the album of its poster and template (the name's digit runs as #, the
extension apart) posted within the window of the album's first file; the
album closes a window after that span and is written as one release
(source album, no file total) named by the members' shared numbers, a lone
member as the release it was. The batch store writes an album's member
names as release_texts kind 6 (member). A counter total of 100 or more under
which one poster started 20 or more releases is a batch counter: those
releases get no file total.
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
edge-bench --group and replay:corpus take --to <article> next to --from, so a
replay covers one window of a group's history while the corpus holds others
(the 2016 ranges next to 2021's and 2025's). deploy/replay-window.sh runs a
window on homelab: every group whose window ranges are all stored is replayed
once into usenet_replay through the voyager-3 app image, 4 groups at a time,
looping while the corpus fill completes the rest.
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
replay:corpus runs edge-bench --group over a group's complete corpus ranges
in article order (one window, carried from range to range, finished after
the last range; edge/src/lease.rs holds the lease JSON the worker and the
bench share) and stores every batch through BatchStore as the worker
would, then reports releases finished, provisional, complete and the bytes
per header of the NVMe and HDD tables. Refuses a database not named
*_replay or *_scratch. The app image carries edge-bench; compose gets a
replay profile for homelab.
Design amended (REBUILD.md 1, 2, 5): the window is carried by one worker a
group, no block leases and no block merger; provisional writes stay for
the lease ends.
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
The indexer finishes files and releases by posting time instead of the
range path: the frontier is the median Date of the last 1,024 plausible
dates (within 7 days of it; a run of 1,024 agreeing dates moves it), and
a release is finished, judged and written once the frontier is 24 h past
its last part, flagged edge when posted within a day of the block's start.
At the block's end the rest is written provisionally as edges for the
block merger. Files short of parts stay in their release (partial: the
store writes complete = false). Gone: fragments and the cut-off,
uploads_of (online now), check.ts and the fixture comparison; CI replays
the fixtures through edge-bench instead. The wire carries finished, edge
and partial.
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
Each release chunk is a release inserted once: release_key the SHA-1 of
the identity, the owner group and the first article (deterministic; no
numbering under a group lock), identity_key the identity's SHA-1 bytes,
poster_id and group_set_id through catalog_refs (cached per process),
source as a code (weak-stem added), the name into release_texts. A
finished release's candidates are claimable at once; a provisional tail
release is finished by a later chunk of its key, a finished one sent again
is left alone. Gone: joins into earlier releases and across groups, key
numbering, the group lock, the fragment merger. Interim: chunks, counts
and samples keyed by ranges.block_id, a range without a block gets one.
Readers follow: search (posters and group_sets joined, words through
release_texts, posters by id, source by code), the par2 merger, the
sidecar folder, the judge's scraped rule, the spot inspector, the spotter,
the NZB service, releases:regroup, edge:bench. Tests insert fixtures
through tests/functional/helpers/releases.ts; the suites pass again.
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
One table and one trigram index for every string search matches against:
the release name (kind 1, stored once more) and what anyone learns later
(the prober's file names inside an archive, NFO titles, par2 set names,
media titles), appended by insert, never an update of the releases row.
Kinds and the per-kind cap in lib/release/text_kind.ts (textsToStore).
The judge's dirty mark follows an insert or delete. Migration 6000 loses
its two trigram indexes on releases.
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
releases rebuilt without its text copies: poster_id, group_set_id,
bytea keys, source smallint, edge, settled_at as the finish. posters by
id, group_sets. blocks as the lease unit, ranges.block_id. fragments,
range_blocks, release_final, the settler and the fragment merger gone;
release_chunks, rule_counts and rule_samples keyed by block. Cold tables
on CHUNK_TABLESPACE. Migration 6000 truncates the catalog and refuses
without VOYAGER3_RESET=1.
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
Eweka stops a compressed XOVER reply at the first ".\r\n" in its own
output (78 01 2e 0d 0a and nothing more), which some small replies begin
with; the text path waited for more until the reply timeout. ranges:dates
uses plain connections of its own instead of the pool.
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
ranges:dates samples each active group's posting dates from the news
server (group_dates) and estimates every range's date (ranges.est_date).
ranges:capture --from --to marks a window's ranges of every group for the
corpus. The corpus fill takes capture ranges in priority order, done ones
too, and remembers complete ones (ranges.stored_at).
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
A 24 h posting-time window per block instead of write-early-merge-later,
albums for counterless series, poster/group-set ids and binary keys in
releases, cold tables on the HDD, the corpus test before the reset.
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
History fetched ahead into the corpus (bulk HDD) is ~60 TB raw for the
1,180B missing articles, and the indexer keeps only ~38% of headers.
With CORPUS_FILTER (default on), Corpus.write leaves out the lines the
edge worker drops on their own, decided per line as RangeIndexer.line
does: a crosspost another indexed group owns, and a subject that names no
file part (~48% of the indexed ranges' headers). Release-level rules
(garbage, per-article, executable, poster) judge files across chunks and
aren't applied. The chunk is compressed again as zlib (the worker takes
it as before); beside it a .drops (gzip JSON) keeps the rule version,
the owners' priorities and per rule the article numbers (as differences,
crossposts by owner), and range.json says filtered. A test shows the
worker makes the same releases, candidates and fragments of a filtered
chunk. corpus:filter filters chunks stored raw (planned capture ranges;
--all for every range).
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Migration 1200 passed regexes with `?` through knex, which made them
bindings: the live function held blu-$1ray, web-$2dl and h\.$326[45], so
BluRay, WEB-DL and H.264 names without a resolution were Misc. 93% of
recent releases were Misc, as the category came from the name alone.
category becomes a plain column (DROP EXPRESSION, no rewrite) set by a
trigger: the name's category (fixed, no `?` anywhere; adds XXX,
audiobooks, console and PC games, and TV codes like S01e07[ and
S04E01E02), else the prober's inner name's, else one given before (a
spot's), else the groups'. The spot inspector sets a spot's category
(spot_category) on Misc releases it links and on its own releases. The
Newznab and web category lists gain Console, PC Games, Audiobook, Anime,
XXX and Comics. releases:recategorize recomputes existing rows in id
batches, writing only changed rows and waiting under a free-space floor.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
59% of settled releases had complete = NULL: without an [i/N] counter
the settler left it unknown, and the "phase 2 reads par2 sets for those"
step was never built. A release is now complete when every file its par2
sets describe is there by name or by its first-16-KiB hash, or, for an
upload renamed after its par2 was made, when it has as many non-par2
files and bytes. Decided when a par2 index is probed, after a par2 merge,
and before the pruner drops a set's file list; the batch store, settler
and merger keep a decided value instead of resetting it to NULL.
releases:par2-complete decides the releases probed before this.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
7z keeps its file list at the end of the archive, so the 7z-tail
candidates (the last segment of the highest volume) were queued but
never probed, and the .7z.001 start only gave the type. The prober now
claims 7z-tail candidates after the archives; lib/sevenzip finds the end
header in the article, decodes the LZMA-packed header (a TypeScript port
of the LZMA SDK's reference decoder) and lists the files with their
sizes. An AES-encrypted header means passworded without names. Names
feed the inner name and the judge's fake rule.
An unsplit .7z of several segments now also gets a tail candidate, and
.zip files an archive candidate: their first segment's local headers
give the first entries and whether they are encrypted. Both in the Rust
edge and the TypeScript indexer.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
The live tail reads an upload in pieces while it is posted, and each
piece queued a 7z-tail candidate for the last volume it had seen so far
(release 306905984: seven, for .002-.008). Only the last volume's end
holds the 7z listing. When candidates are stored, and when releases are
merged, waiting tails below the release's highest volume are deleted.
All tails of the highest volume stay: a piece that saw only part of it
queued a segment that is not its last. RAR is unaffected: it lists its
files in the first volume's header and gets no tail candidate.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Eweka drops quiet connections, so the next command on one failed
("Spot inspection failed: NNTP connection closed"), and idle connections
held account slots that downloads could use. The coordinator pool now
closes free connections (and the plain one) on its sync once idle for
60 s; dry prober lanes, the spotter between rounds and the spot inspector
with nothing to inspect close theirs too, reconnecting when work comes.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Most par2 sets describe readable uploads whose files are never hashed, so they
were never complete and kept forever (19 GB). Candidates added to a release
after it settled were never marked ready (93,642 since 2026-10-13); the batch
store and merge mark them now, migration 5100 repairs the waiting ones
(LOG.md 2026-10-20 (87)).
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
A round behind by two days of the probe backlog read 3.5M sets and never ended;
a par2 file shared by up to 190k files loaded every release it matched and the
merger ran out of heap every few minutes (LOG.md 2026-10-20 (85)).
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
A full round took 35-38 min of each hour and grew with every inspection
the prober added. Triggers note changed releases in judge_dirty; the
judge judges those (and releases a file hash of theirs tipped to
repeated) by every rule, and every release in a daily full round for
what time changes.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
"Murdoch Mysteries S8D1 [00/80] - \"….nzb\"" beside its "[01/94]" upload
was a release of its own (23k sidecar-only releases on the corpus, 15k with
the poster's bigger release within 10 minutes). The settler queues small
settled releases (fold_candidates); the merger's third pass folds one whose
files are all sidecars into the poster's bigger release in the same group
within 10 minutes: the one sharing a 4+ letter word with its subject, else
the one under the same counter total, else none (via 'sidecar').
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Voyager 1's second filter (D42) for voyager 2. Hourly, in order (the first
rule wins): dead (every probed first article gone), fake (a film, episode or
music name whose content is a program: archive entries all executables, or
the file itself by name or sniffed exe), repeated (the same first 16 KiB and
size in 20+ releases), locked (passworded, no password known), unusable
(random names, no inner name, no merge, no spot, a day after settling).
Verdicts in release_verdicts (rule, version) hide releases from search
(judged shows them); a verdict whose rule no longer matches is taken back.
Nothing dropped yet. Service judge; releases:judge --once --dry counts.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Two processes, one NNTP connection each (the coordinator 40 -> 38):
- spots:read (service spotter): free.pt's overviews after the last article
seen, each spot's full header verified (lib/spotnet from voyager 1, as
Spotweb), its description, website and NZB articles; the password its
title or text names; delete/dispose messages applied.
- spots:inspect (service spot-inspector): each spot's compressed NZB read,
its files listed (spot_files), each matched to the header pass's release
by the same release identity in one of its groups within 48 h
(release_spots); unmatched spots matched again hourly, after 24 h a
release of their own (source spot) with the spot's NZB.
- release_passwords: per release and password, its source (nzb, spot-title,
spot-text) and spot; the catalog keeps only passworded.
XoverClient.head(); PROBE_CONNECTIONS default 10.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
The prober stores, per probed article that starts a file, the MD5 of its
first 16 KiB (file_hashes), and per par2 body the set's whole file list
(par2_files, par2_sets): FileDesc holds the same hash per file, whatever
the file was renamed to. releases:merge (the merger service, no NNTP)
merges every minute the releases of one group, posted within 48 hours of
a release holding the set's par2, whose files the set describes, into the
one with the most files: chunks, probe candidates, inspections, hashes and
counts move; the par2's shared name when none was readable; release_merges
keeps what went where. Obfuscated uploads whose files were each a release
of their own ("[9/20] - "14f5999a…"", hep: 13,232) become one.
Also: corpus_sweep.py counts the marked ranges without psql's command tag.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Ranges marked `capture` keep their XOVER chunks in CORPUS_DIR
(corpus/<group>/<start>-<end>/, edge-bench's layout, range.json once
complete; only chunks that inflate). The prefetcher reads a stored chunk
instead of fetching it, so a reset re-indexes them without the news
server; a chunk a worker reports corrupt goes from the corpus too.
edge-bench gets --releases (release chunks without payload, with their
first subjects) and ships in the edge image; deploy/corpus_sweep.py marks
more ranges (capture), replays the corpus with the deployed build (replay)
and reports suspects and the change since the previous build (report).
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
User: Rust is the product (option a). tools/edge-bench-rs moves to edge/
(crate edge; binaries edge-node, the worker, and edge-bench, the replay).
CI's new job edge runs cargo test and edge/check-fixtures.sh: both
RangeIndexers on three committed captures (edge/fixtures, 25k overview
lines each of a.b.models, a.b.ath and a.b.boneless in 5 chunks, 3.5 MB),
compared field by field with check.ts; deploy waits for it. The
TypeScript RangeIndexer stays as check.ts's reference. README: what it
is now; the benchmark findings as history.
Random names: looksRandom missed long tokens of capitals and digits
("51294P1CUKR2R35X", release 74539068, readable and a Newznab title).
12+ capitals and digits switching 6+ times between them are random
(COD4MW2REMASTERED switches 4 times, BEATLES1964REMASTER twice); both
implementations, identical on the fixtures.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
From "draft for approval" to the concept of voyager 2 (voyager 1's
CONCEPT.md and D1-D42 are superseded). Phase 2: what's built (the
prober), and voyager 1's features not yet here: spots, NZB import,
passwords, SFV, RAR walks, 7z listings, body naming of obfuscated
subjects, after-probe garbage drops; where its exported data is.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Eweka stopped answering on all 40 of the coordinator's long-lived
connections (quiet sockets; a fresh connection fetched the same chunk in
20 s): with a 10-minute timeout the prefetcher fetched one chunk a
minute. The timeout (silence, not duration) is 2 minutes now: 365-411k
articles/s after the restart. tools/probe_chunk.ts times a chunk
compressed and plain.
README and REBUILD.md: edge-node, the Rust worker, is the one running
(8x less CPU per header, 20x less memory than edge:node in production).
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
For workers on a shared line (poweredge-kantoor: 160 Mbit/s): the range
stream is read at most at the cap, TCP holds the coordinator back. The
cap is the node's, split over its processes (also EDGE_MAX_MBIT).
REBUILD.md: where workers run (connections set the speed; remote workers
take CPU off homelab).
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
