jordan/tidaldb
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity
2c25b6d375
tidaldb/docs/planning/milestone-6/phase-6.md
jordan fe8d0c87e7 harden: restore CI verification, remove four wire-level fabrications, instrument the 401 path 
Implements tmp/tidaldb-fleet-hardening (20 planned tasks + 2 found by measurement).

Ring 0 — restore verification. .woodpecker.yaml step pods ran at the namespace
default of 1500m/2Gi, which OOMKilled a prior pipeline and starved the release
gate past its budget. Both push-path steps now declare
backend_options.kubernetes.resources as two YAML anchors declared once on their
first consuming step. The values are CALIBRATED against measured free node
capacity, not against the LimitRange max: `requests: cpu 2` (this roadmap's
original figure) fits on NO node and would sit Pending forever, because
`ci-build-bounds` grants permission and the nodes supply capacity, and those are
not the same thing.

The `nightly` cron described in this file for 216 days was never created, so
tier-3 chaos, the fault classes, mTLS and the PITR test produced exactly zero
signal while reading like standing coverage. nightly-chaos and
nightly-security-ops now alias the anchors and have budgets matching the gate
(their 120/90 were TIGHTER on the same runner, so they would have failed
nightly for a budget reason, not a correctness one). nightly-soak is REMOVED,
not scheduled: it drives 1000 rps for 600s gating on p99 <= 250ms, and the best
node has 1700m free CPU, so it would fail on starvation rather than regression —
manufacturing a nightly false alarm. Its commands move verbatim to
docs/runbooks/nightly-soak.md.

Ring 1 — four fabrications removed from the wire.
- scatter_merge sorted and truncated without re-stamping rank, so /feed and
  /search returned 1,1,2 under full placement. Reuses merge_cross_shard's
  existing stamp; asserted on BOTH the multi-group merge path and the
  single-group [only] fast path that bypasses it.
- aggregate_region_row's None arm invented `applied_events: 0` plus a deficit
  derived from it. applied_events/lag_events are now Option<u64>, null on the
  wire. leader_last_seq was also unwrap_or(0), so a node that could not reach
  the LEADER computed 0 - applied = 0 for every region and reported a converged
  cluster it had never measured — a fabrication pointing the dangerous way.
- tidalctl inferred NO REPORT from `applied == 0 && lag > 0`. That heuristic was
  actively hiding the PVC-wipe shape: a measured zero with a real deficit
  rendered as "no report" instead of BEHIND. Now read off the wire; converged
  exits 0, partitioned still exits nonzero.
- /sharded/* answered 201/204 for single-copy writes with nothing anywhere
  saying so. Now requires `x-tidal-ack: local`, rejecting with 400 via the
  existing invalid_input path. Six call sites migrated, not the two this
  roadmap predicted — including docs/runbooks/cluster.md §16.3, which told
  operators to run a quorum-write probe via POST /sharded/items. That probe
  cannot verify quorum: the surface applies locally with no WAL append. It was
  used as the safety check between every step of a staged deploy earlier today.

Ring 2 — observability. JSON_LOGS was already implemented and the deployment
simply never asked for it; the StatefulSet now sets it, plus
TIDAL_SERVICE_NAME=tidaldb because enabling it silently renames the
VictoriaLogs `service` stream field and would have blinded every query keyed on
it. Adds tidaldb_usearch_replicated_vectors_total, incremented on BOTH the
origin (wal_blob_first -> Ok(Some)) and the follower apply path — counting only
the origin would mean each vector lands on exactly one node, replicas never
agree, and the alert built on it pages forever.

Found by measurement, not planned: the 401 path discarded every fact about
every rejection. Traefik has served 101,858 rejected requests to the public
ingress — 87.6% of all its traffic — with no record of who or why anywhere.
unauthorized_response now emits reason (missing_token vs invalid_token, the
distinction that separates a scanner from a rotation that missed a consumer)
and the forwarded client. The token is never logged.

Also: scripts/restore-fleet.sh --cluster started the soak monitor while
deliberately leaving its gate suspended, orphaning a watcher that has reported
"0/30 green nights" for 13 days. The pair now moves together. Doc-guard's
three-warning backlog is cleared with real backfill for M4/M6/M12.

Verified: fmt clean; clippy 5 crates 0 new warnings (74 vs 74 baseline,
counted in a detached worktree at HEAD); lib 2110 passed; cluster_sharding 5;
cluster_runbook 10; tidalctl 38; doc-guard 0 warnings. Playwright 32/34 with
the two remaining failures asserting the rank fix against the not-yet-rolled
image — they are the post-deploy proof.
2026-08-30 20:55:58 -06:00

3.4 KiB
Raw Blame History

m6p6 — Notification Capping + Adaptive Preferences + Creator Profile Modes + M6 UAT (✅ COMPLETE 2026-02-23)

Phase spec and acceptance criteria: ROADMAP · Milestone 6 · Phase 6. Milestone index: README.md. Backfilled record.

What shipped

  1. Notification capping (tidal/src/db/notification_tracker.rs). NotificationCaps { max_per_creator_per_day, max_total_per_day } attaches to a query via RetrieveBuilder::notification_caps(caps) and is enforced as a post-diversity pass, with per-(user, creator, date) delivery counts tracked so the cap spans queries rather than only trimming one result page.

  2. Adaptive preference learning rate (tidal/src/entities/preference.rs). The EMA alpha decays logarithmically with a user's update count:

    alpha = base_alpha / (1 + ln(update_count + 1))      // base_alpha default 0.1

    At count = 0 this is exactly base_alpha. Early signals move a cold user's vector hard; later signals refine it gently, which is what keeps a well-established taste profile from being yanked by one outlier interaction. Unit coverage for the decay curve lives beside the implementation (preference.rs tests, incl. the explicit count=0 ⇒ alpha=0.1 case).

  3. Creator profile modes. RetrieveBuilder::for_creator(creator_id) adds the creator filter and restricts candidate generation to that creator's items, so for_creator(x) + for_you ranks x's catalogue by the querying user's preferences while for_creator(x) + hot ranks x's catalogue by heat.

  4. The M6 UAT (tidal/tests/m6_uat.rs) — 9 #[test] functions over a shared fixture, exercising the full use-case surface.

Evidence

  • tidal/tests/m6_uat.rs (9) and tidal/tests/m6p6_creator_profile.rs (6).
  • Recorded at close in the ROADMAP status row: 1,082 total (835 lib + 247 integration), 9 m6_uat passing; the same row records that all prior milestone UATs (m2, m3, m4, m5, m5p4) continued to pass — the phase's "no regression" criterion.

Divergence from the plan

The "update counts are not persisted" limitation is stale — and it was closed by M12, not M7. tidal/src/entities/preference.rs:37-45 still carries a # Known Limitation block saying update_counts is in-memory only, that every user resets to count = 0 on restart, and that persisting it is "deferred to M7 (Production Hardening)". Both halves are now wrong:

  • The count is persisted. The multi-vector preference checkpoint format serializes it — [update_count:8 LE][importance_at_anchor:4 LE][anchor_ts:8 LE] (tidal/src/entities/multi_preference.rs:127,149) — with MultiPreferenceVectors::checkpoint / restore (multi_preference.rs:642,703) writing and reading it, and PreferenceVectors::insert_restored (preference.rs:239) loading a legacy single-vector row back as a cold-start user with its count intact. Round-trip and legacy-row tests exist (checkpoint_restore_roundtrip_multi_cluster, restore_reads_legacy_single_vector_rows_as_cold_start).
  • The milestone that closed it was M12 (multi-vector user preference modelling), not M7. See milestone-12/multi-vector-preference.md.

The stale doc comment is in Rust source and is out of scope for this backfill; recorded here so the next reader of that comment does not re-plan work that is already done.