- Replace Process.sleep(N) with explicit synchronization:
render(view) to force LV mailbox drain, PubSub assert_receive for
leader election, and render_hook sync guarantees
- Add batch_packet_fixtures/2 using Repo.insert_all for bulk test
fixtures (200 inserts -> 1 round trip)
- Remove sleeps from historical_loading_test — tests query DB
directly, don't need LV async processing
Coalesce rapid PubSub notifications into single 2-second delayed DB query. Remove tests for removed :update_time_display handler.
Ultraworked with Sisyphus
Co-authored-by: Sisyphus <clio-agent@sisyphuslabs.ai>
Replace full-list assign with stream_insert + phx-update=stream. Add DOM IDs to all packet rows. Skip deprecated phx-update=prepend.
Ultraworked with Sisyphus
Co-authored-by: Sisyphus <clio-agent@sisyphuslabs.ai>
Bound batch DISTINCT ON queries by input count. Guard Repo.all(Devices) with exists? check. Add trigram GIN index on base_callsign and (has_weather, received_at DESC) partial index.
Ultraworked with Sisyphus
Co-authored-by: Sisyphus <clio-agent@sisyphuslabs.ai>
Remove redundant in-memory sort (DB already orders), shortcut get_packet_field when field found at top level, use pre-computed has_weather boolean instead of field-by-field scan
Ultraworked with Sisyphus
Co-authored-by: Sisyphus <clio-agent@sisyphuslabs.ai>
Remove unused 30s timer, skip heatmap recluster on every single packet, trail dedup O(n^2)->O(n)
Ultraworked with Sisyphus
Co-authored-by: Sisyphus <clio-agent@sisyphuslabs.ai>
The unstyled <h1>Packets</h1> template was dead code - no route in the
router renders it (/packets uses PacketsLive.Index LiveView instead).
Removed the template and its test. Added findings.md with all audit
results from comprehensive security and code quality review.
- k8s: 2 replicas, POOL_SIZE 25, cpu 2000m, mem 1Gi (was 1 replica, 5, 500m, 512Mi)
- telemetry poller 10s -> 60s; drop pg_database_size, pg_stat_statements,
pg_table_size and pg_indexes_size (these were ~7100s of cumulative DB time
on their own)
- get_latest_packet_for_callsign: bound by :packet_retention_days so partition
pruning kicks in
- get_nearby_stations_knn: ST_DWithin spatial pre-filter (default 500km) so
the GiST geography index cuts candidates before DISTINCT ON sort
(EXPLAIN: 5793ms -> 400ms)
- New migration: partial functional indexes on upper(object_name) and
upper(item_name), built per-partition CONCURRENTLY then attached to parent.
Lets get_latest_packet_for_callsign use BitmapOr across all 3 upper()
indexes instead of falling back to a scan
When :beam_lib.chunks fails (e.g. a project module hasn't been loaded
into :code.which yet), the existing with returned the raw error tuple
as the 'line' value, which then crashed format/1 via String.Chars.
Add an explicit else clause so the function always returns nil or an
integer.
- Split Analyzer.analyze/1 into compute_state/1 (slow, scans all BEAMs)
and a fast filter step. analyze/1 now accepts :precomputed_state opt.
- Mix.Tasks.Compile.Unused.run/1 picks up precomputed state from process
dict :mix_unused_extra_analyze_opts when present (test escape hatch).
- AnalyzerTest setup_all computes state once; exclude tests pass it
through analyze/1, dropping per-test cost from ~700ms to <1ms.
- UnusedTest setup_all does the same — the consolidated severity-args
test drops from 3.4s to ~5ms.
- Trim further sleeps in HistoricalLoadingTest (100->30ms) and
MovementTest (50->20ms, receive timeouts 100->30ms).
Total suite: 40.6s -> 37.7s; 2488 tests, 0 failures.
- packet_replay: stop_replay test reduces fake-task receive timeout from 5s to 200ms
- broadcast_task_supervisor: scheduler_usage takes configurable sample seconds;
test uses sample-pair API (instant) instead of blocking 1-second sample
- health_check: ask_if_alive timeout is now configurable; unresponsive-pid test
uses 50ms timeout instead of waiting full 5-second default
- mix_unused/analyzer_test: cache analyze() result in setup_all so 4 tests share
one analyze pass instead of running 4 separate scans
- mix/tasks/compile/unused_test: consolidate three slow severity tests into one
- log_sanitizer + packet_field_whitelist: cap property-test runs at 25 (was 100)
- historical_loading: trim 200ms post-event sleeps to 50ms
- movement: refute_push_event timeout from 200ms to 50ms
Total suite time: ~45s -> 40.6s; 2488 tests, 0 failures.
- Make LeaderElection check_interval, max_cluster_wait, and cluster_check_interval
configurable via Application env so tests can override with short values
- Slow LeaderElection tests now use 50-150ms timeouts instead of 3-6s
- Remove :slow tag from unused_test (compile cache makes reruns cheap)
- Drop :slow from test_helper exclude list
All 2490 tests now run in ~45s (up from 2485 with 5 excluded).
Adds a self-contained 'unused' Mix compiler task that finds public
functions in this project no caller in the project ever invokes:
- MixUnused.Analyzer: enumerates public functions across the project's
compiled BEAM files (filtered by app), walks the abstract code in
each BEAM via :beam_lib to collect every remote {Module, fun, arity}
call observed at compile time, and subtracts the call set from the
def set. Exposes :exclude (modules or MFA triples), :compile_path,
and :app options. Behaviour callbacks (@impl true and any listed
in behaviour_info(:callbacks)) and well-known entry points
(start_link, child_spec, mount, render, GenServer callbacks,
__info__, module_info, etc.) are excluded automatically.
- Mix.Tasks.Compile.Unused: a Mix.Task.Compiler that compiles the
project then asks the analyzer for unused defs. Severity is
configurable via the :unused project key; per-file exclusion of
test/ paths is on by default.
Inspired by https://github.com/hauleth/mix_unused — uses BEAM
introspection rather than a compilation tracer to avoid the
chicken-and-egg problem of recompiling the tracer module itself
during analysis.
