dragonflight/services/mam-api/src
ZGaetano 4be12c6f9a fix(stability): spread capture encodes across all GPUs + GOP parse + filmstrip retry
VIDEO FREEZE UNDER BURN (transient stall, self-recovers): all 8 capture
sidecars ran NVIDIA_VISIBLE_DEVICES=all with no -gpu selector, so ffmpeg
nvenc put every session (8 HEVC masters + 8 HLS = 16) on physical GPU 0
while the other two L4s sat idle. GPU 0 NVENC hit 86%, encode fell below
realtime, the framecache ring lapped → video froze → caught up → recovered.
Bridge verified smooth at 60fps throughout. FIX: node-agent now round-robins
each sidecar to a GPU by capture port (port % detected-GPU-count) via
NVIDIA_VISIBLE_DEVICES, honoring an explicit gpuUuid when set. Auto-detects
GPU count from nvidia-smi (override CAPTURE_GPU_COUNT). ~3 encoders/GPU now.

GOP PARSE: Number.parseFloat('60000/1001') returns 60000, making GOP 120000
(near open-GOP) instead of ~120. Added parseFps() to handle rational rates;
fixed hevcNvencArgs + buildHlsVideoArgs.

FILMSTRIP: RustFS object store intermittently returns NoSuchKey on GET for
keys that List/Head confirm exist, blanking the strip. Generation/queue/DB
all verified healthy (13/15 assets HAVE filmstrips). FIX: API now serves the
filmstrip JSON through itself with retry-on-NoSuchKey (succeeds within a
couple attempts) instead of handing the browser a signed URL — also closes
the S3 CORS gap. Frontend updated to consume the direct JSON.
2026-06-04 15:56:44 +00:00
..
ampp feat: AMPP folder sync integration — pre-create folder hierarchy on upload, expose lookup endpoint for Script Task: client.js 2026-04-18 13:42:07 -04:00
auth fix(audit): critical security hardening and ops reliability fixes 2026-06-03 10:42:57 +00:00
db feat(recorders): hardware-identity model with Enable/Disable lifecycle 2026-06-04 03:14:43 +00:00
middleware fix(playout): repair failover, authenticate scheduler self-calls, fix playlist walk + CasparCG consumer syntax 2026-05-30 14:51:35 +00:00
routes fix(stability): spread capture encodes across all GPUs + GOP parse + filmstrip retry 2026-06-04 15:56:44 +00:00
s3 fix(s3): keep-alive agents + long timeouts to end socket starvation 2026-06-04 12:53:28 +00:00
tasks Revert "auth: top-to-bottom rework — local accounts, RBAC + client tag, audit log, env-bootstrap" 2026-05-27 03:28:05 +00:00
index.js fix(mam-api): selfHeartbeat writes last_seen_at so primary node isn't stale-failover-killed 2026-05-30 18:57:20 -04:00
scheduler.js fix(scheduler): orphan grace window must use recorder.updated_at not asset.updated_at — asset is created at recording START not STOP 2026-06-03 14:03:32 +00:00