From the studio a single connection to Hetzner nbg1 (188 ms away) is
capped near 0.1 MB/s, although a browser speed test shows 50 Mbps.
Measured: 1 connection 0.13 MB/s, 8 -> 0.64, 24 -> 1.69 MB/s. At the old
speed a ProRes MOV could not finish inside the 30 min request timeout.
- Worker 1.7.2 uploads files >= 16 MB as S3 multipart: 8 MB parts on
uploadConcurrency connections (config.json, default 16), each part
retried up to 4 times; aborts the upload if a part keeps failing.
- artifact-presign returns uploadId + per-part presigned URLs when asked
for multipart; new E20b artifact-complete stitches or aborts.
- Old workers and old servers fall back to the single PUT.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Preview jobs looked stuck after aerender finished: the worker was silently
uploading the review MP4s and the ProRes MOV over a slow uplink, and a
restart mid-upload threw the preview away.
- Worker 1.7.1 replaces the job's log tail with live upload status and logs
each upload's size, duration and speed.
- The ProRes MOV now uploads after finalize via a persistent queue
(HighResUploader, resumes after restart, 5 attempts) and attaches through
the new E13b endpoint POST /api/ext/render/jobs/{id}/attach-highres.
A MOV from a superseded preview is discarded.
Deploy the server before the 1.7.1 worker.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
A FAILED delivery item never sets fileCount/totalBytes, and the worker's
Dictionary-based JSON payload serializes those as explicit nulls (the
WhenWritingNull option only suppresses nulls on POCO properties, not
dictionary values). The server's schema allowed the fields to be omitted
but not null, so the report was permanently rejected with 422 - and the
worker only ever treated 409/404 as terminal, so it retried the same
malformed payload forever across restarts.
Blender work is not a shot submission: no Export, no version, no
validation, QC or delivery. It is "Ctrl+F12, but on the farm", so it is
modelled as what it is — a queue entry and nothing more.
That turned out to need very little. Claim, progress, fail and
complete-render already tolerated a job with no Export behind them, so a
Blender job reuses the whole lifecycle; RenderJob gains projectId, label
and submittedBy so a job with no shot code still has something to show.
Capability gating is new and was already missing: the claim now filters
job types by the machine's reported tools. Without it a workstation
without Blender would claim Blender work and fail it repeatedly, burning
the attempts while a machine that could have rendered it sat idle. The
worker asks for Blender jobs unconditionally and the server refuses them
— trusting the worker to self-select would have left the same hole.
The runner treats a clean exit with no saved frames as a failure:
Blender exits 0 when a frame range is empty or an output path is
unwritable, and a job that silently rendered nothing is the worst
outcome available.
Standalone jobs auto-requeue on a transient failure like any other work;
delivery builds are deliberately excluded, since a half-written package
wants a human first.
Adds the Blender add-on (BlenderPanel/) and its docs. The add-on refuses
to submit an unsaved file — the farm opens the .blend from disk, so
unsaved changes would render the previous state without warning. Its
payload is pinned by a contract test, since add-on and server only ever
meet over JSON.
Worker 1.5.0. Adds migration 20260815000000_blender_render_jobs.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
"What is running where" was answerable only by asking someone. Each tool
now reports its own version on startup into a ToolVersion registry, shown
on Pipeline → Machines.
The worker's version moves out of config.json and into the assembly:
a hand-maintained string on each machine is exactly the thing that goes
stale, and a machine reporting a version it is not running is worse than
no version at all. publish.ps1 stamps the git SHA via SourceRevisionId,
so a deployed exe traces back to a commit (1.1.0+feb1c7b).
The panel gains PANEL_VERSION, shown in its title and every diagnostics
report, and pings on load.
Staleness is derived rather than declared: the newest version seen for a
tool is the benchmark, so updating one machine is what flags the rest.
There is no expected-version list to keep in sync — which would have been
one more thing to forget.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Completes RenderPipeline2 phases 3-5. Additive throughout: with
validation.enabled=false the pipeline behaves exactly as before.
Validation (§8) runs on the worker straight after a render, while it
still holds the job and the files are local. Everything is read from the
EXR header — one 64 KB read per sampled frame, no pixel decode — so the
engine needs no oiiotool. The render job deliberately stays open through
the stage, so a worker that dies there is reaped and retried instead of
stranding the export in VALIDATING.
Timecode is the primary check and is anchored on the plate, never the
API: measured on the live show, Shot.timecodeStart is the editorial
in-point and sits one handle-length (8f) ahead of the plate, so checking
against it would pass the defective exports and fail the correct ones.
Reference precedence is manifest timing plate -> plate found on disk ->
API value, and the API rank is WARN-only. The panel now resolves plate
provenance by walking the comp's own footage layers, so the worker never
guesses which element carries the comp's timing.
Verified against real studio renders via the new --validate mode: it
catches both defect classes seen in shipped work — the +8 handle offset
and frame-derived timecode (00:00:41:17 = frame 1001 at 24fps).
QC (§10) is one service behind two front doors, API-key for the panel
and session for the web. PASS means ready to upload and nothing else;
FAIL sets the comp task to CHANGES. The panel gains a QC section that
imports the real EXR sequence at full res rather than the H.264.
Delivery (§11) rides the existing queue as a DELIVERY_BUILD job: hard
links with a copy fallback, verification by re-hashing delivered bytes
against the validation-time digest, manifest.json and the CSV report.
fsutil hardlink list is unavailable over SMB, so link identity is never
queried — reading the bytes back covers links and copies alike.
Also fixes ApiClient.ProgressAsync throwing on transport failure, which
would have aborted work in flight during a brief server outage; progress
reporting is best-effort by contract.
Removes the stale RenderWorker/ rule from the root .gitignore. The
worker's source is tracked like the rest of the repo and its own
.gitignore already covers bin/, obj/ and publish/; the rule silently
excluded new files while the existing 20 stayed tracked.
Not yet applied: migration 20260813000000 was hand-written because the
dev database was unavailable — run prisma migrate deploy and verify
before relying on it. None of this has run against a live server.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
One "Queue Export" click now renders the EXR sequence, then rebuilds the shot
headlessly with the studio slate/overlay template to produce the delivery MOV
and review MP4. Implements RenderPipeline2 phases 1-2 plus the preview stage.
Server:
- New models Export, RenderJob, ExportEvent, Machine, WorkerHeartbeat, plus
Project.deliveryConfig and per-submission slate fields (Export.vfxScope,
Export.submissionNote, inherited from the shot's previous export).
Both migrations are purely additive; no existing column is touched.
- lib/render-pipeline: server-enforced state machine, transactional version
increment with supersede, atomic FOR UPDATE SKIP LOCKED claim gated by
machine availability windows, and a lease reaper run from instrumentation.ts.
- /api/ext/* endpoints for the panel and workers; session-auth mirrors under
/api/render and /api/machines for the web UI.
- Pipeline pages: render queue, export detail, machine monitoring, plus an
Exports tab on shot detail.
RenderWorker (.NET 8 Windows service, new):
- Registration, heartbeat as cancel channel, claim loop, aerender runner with
progress parsing and stall watchdog, crash recovery and disk-spooled
reporting that survives server downtime.
- Preview stage: headless AE assembles the preview comp into a throwaway AEP
with both output modules queued, then a single aerender pass renders them.
Preview jobs are not claimed while an interactive AE session is open, so an
artist's project is never taken over.
AE panel: Queue Export with live status polling, urgent flag, retry, and the
VFX Scope / Submission Note fields. Every existing panel action is unchanged.
Preview chaining ships disabled behind SystemConfig preview.enabled.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
twotalesanimation