Matthew Meszaros 42ddb78e60 feat: make the realtime service survive a client that spams channel joins, and stop both clients from being that client: ws_join is now spent per phx_join in a shared RealtimeWeb.ChannelGuard that every channel runs BEFORE its Auth lookup (it was only ever spent on the socket handshake, so an established socket could issue unlimited joins and each one cost a Postgres query), handshakes move to their own ws_connect bucket so a reconnect storm cannot eat the budget a client needs to rejoin with, the retry_after_ms hint now points at the next window instead of shrinking as overage grew, every channel's join rejection carries the numeric code the API reference already promised (plus 4005 for a malformed topic), BulkChannel only pushes an event to the user who owns the operation instead of to anyone who guesses the id, and the web and admin clients treat a rate_limited join as transient (wait out the hint, keep the socket) while backing off per-channel rejoins on a decaying schedule rather than reconnecting on the 120ms floor forever; adds an ExUnit suite for realtime and a vitest suite for admin, both wired into CI

2026-08-26 21:14:34 -07:00

..

pubsub

feat: bridge realtime events over Redis pub/sub when Google Pub/Sub is unconfigured (RedisBus publisher in backend/consumer plus Realtime.Redis.EventSubscriber and a shared EventBroadcaster) so dashboard live updates and presence-driven collaboration actually fire in local dev

2026-06-12 04:49:58 +02:00

redis

feat: bridge realtime events over Redis pub/sub when Google Pub/Sub is unconfigured (RedisBus publisher in backend/consumer plus Realtime.Redis.EventSubscriber and a shared EventBroadcaster) so dashboard live updates and presence-driven collaboration actually fire in local dev

2026-06-12 04:49:58 +02:00

api_key.ex

Add local Sentry logging

2026-02-14 05:49:49 +01:00

application.ex

feat: make the realtime service survive a client that spams channel joins, and stop both clients from being that client: ws_join is now spent per phx_join in a shared RealtimeWeb.ChannelGuard that every channel runs BEFORE its Auth lookup (it was only ever spent on the socket handshake, so an established socket could issue unlimited joins and each one cost a Postgres query), handshakes move to their own ws_connect bucket so a reconnect storm cannot eat the budget a client needs to rejoin with, the retry_after_ms hint now points at the next window instead of shrinking as overage grew, every channel's join rejection carries the numeric code the API reference already promised (plus 4005 for a malformed topic), BulkChannel only pushes an event to the user who owns the operation instead of to anyone who guesses the id, and the web and admin clients treat a rate_limited join as transient (wait out the hint, keep the socket) while backing off per-channel rejoins on a decaying schedule rather than reconnecting on the 120ms floor forever; adds an ExUnit suite for realtime and a vitest suite for admin, both wired into CI

2026-08-26 21:14:34 -07:00

auth.ex

feat: make the realtime service survive a client that spams channel joins, and stop both clients from being that client: ws_join is now spent per phx_join in a shared RealtimeWeb.ChannelGuard that every channel runs BEFORE its Auth lookup (it was only ever spent on the socket handshake, so an established socket could issue unlimited joins and each one cost a Postgres query), handshakes move to their own ws_connect bucket so a reconnect storm cannot eat the budget a client needs to rejoin with, the retry_after_ms hint now points at the next window instead of shrinking as overage grew, every channel's join rejection carries the numeric code the API reference already promised (plus 4005 for a malformed topic), BulkChannel only pushes an event to the user who owns the operation instead of to anyone who guesses the id, and the web and admin clients treat a rate_limited join as transient (wait out the hint, keep the socket) while backing off per-channel rejoins on a decaying schedule rather than reconnecting on the 120ms floor forever; adds an ExUnit suite for realtime and a vitest suite for admin, both wired into CI

2026-08-26 21:14:34 -07:00

connections.ex

feat: keep the per-user realtime connection cap at 10 but reap dead connections via a periodic liveness sweep and process pids, so the count reflects real presence and stale sockets never hold a slot

2026-06-12 07:46:13 +02:00

error_reporter.ex

Add local Sentry logging

2026-02-14 05:49:49 +01:00

event_broadcaster.ex

feat: realtime admin dashboard + diagnosability - new Elixir admin:platform channel (JWT-only, gated on users.admin_permissions) mirrors every platform event, the admin app connects through a /getaway-tokenized Phoenix socket client with a throttled react-query invalidation spine, a Live Events firehose page streams the platform with filters/pause/JSON expand, a System Status page runs parallel infra probes (postgres/redis/kafka/schema-registry/realtime/tracking via /admin/system/status), and the worker logs viewer gains follow mode, line-count selection, and copy

2026-07-12 09:00:10 +02:00

event_log.ex

feat: make the org realtime channel a resumable gateway — assign a monotonic per-org sequence + buffer each event in a capped Redis stream (Realtime.EventLog) before broadcast, advertise the current seq in the HELLO and stamp it on every event, and replay the missed gap (re-filtered) on reconnect via resume.last_seq or signal a full resync when the buffer no longer covers the client

2026-06-13 11:23:07 +02:00

oauth_token.ex

feat: authenticate the realtime WebSocket with OAuth access tokens (wmat_ with the realtime scope), validated against oauth_access_grants alongside wmbly_ API keys and JWTs, and map not-a-member/forbidden rejections to close code 4010

2026-06-14 11:02:12 +02:00

rate_limiter.ex

feat: make the realtime service survive a client that spams channel joins, and stop both clients from being that client: ws_join is now spent per phx_join in a shared RealtimeWeb.ChannelGuard that every channel runs BEFORE its Auth lookup (it was only ever spent on the socket handshake, so an established socket could issue unlimited joins and each one cost a Postgres query), handshakes move to their own ws_connect bucket so a reconnect storm cannot eat the budget a client needs to rejoin with, the retry_after_ms hint now points at the next window instead of shrinking as overage grew, every channel's join rejection carries the numeric code the API reference already promised (plus 4005 for a malformed topic), BulkChannel only pushes an event to the user who owns the operation instead of to anyone who guesses the id, and the web and admin clients treat a rate_limited join as transient (wait out the hint, keep the socket) while backing off per-channel rejoins on a decaying schedule rather than reconnecting on the 120ms floor forever; adds an ExUnit suite for realtime and a vitest suite for admin, both wired into CI

2026-08-26 21:14:34 -07:00

redis.ex

Organization, Subscription, Inqueries, limits.

2026-01-27 05:55:48 +01:00

repo.ex

Organization, Subscription, Inqueries, limits.

2026-01-27 05:55:48 +01:00

sentry_finch_client.ex

fix(security): move realtime sentry onto a Finch HTTP client and drop the hackney dependency to clear CVE-2026-47071

2026-06-28 06:18:33 +00:00

sequencer.ex

feat: guarantee in-order gateway delivery via a per-org sequencer pool — serialize each org's events (partitioned by org id) so seq order always equals broadcast order even under concurrent ingest, closing the resume race where a higher seq could arrive first and a disconnect would skip the lower one

2026-06-13 11:31:07 +02:00

subscription.ex

feat: add campaign workflow backend support

2026-06-06 07:49:52 +02:00