diag(merge): log 404 cause + pin Job queue_name + max_jobs=1

backend/src/api/routes/tasks.py

@@ -771,19 +771,43 @@ async def get_merge_job(
         # needed.
         info = await JobQueue.get_job_info(merge_job_id)
         if info is None:
+            # Distinguishes "Redis doesn't know this job_id at all" from
+            # the function/args mismatch branches below. Each branch
+            # returns the same opaque 404 to callers to avoid leaking
+            # the existence (or shape) of other tasks' jobs, but the
+            # server log line tells us which case fired so we can debug
+            # disappearing-job reports (worker on different Redis, key
+            # eviction, race, etc.).
+            logger.warning(
+                "Merge job not found: info=None merge_job_id=%s task_id=%s",
+                merge_job_id,
+                task_id,
+            )
             raise HTTPException(
                 status_code=404, detail=f"Merge job {merge_job_id} not found"
             )
         if info.function != "merge_clips_job":
             # Wrong fn => not a merge job at all; treat as not-found rather
             # than leak which functions exist.
+            logger.warning(
+                "Merge job not found: function mismatch merge_job_id=%s task_id=%s function=%s",
+                merge_job_id,
+                task_id,
+                info.function,
+            )
             raise HTTPException(
                 status_code=404, detail=f"Merge job {merge_job_id} not found"
             )
         # args === (task_id, clip_ids) per merge_clips_job signature.
         # Mismatch means the caller is asking about a job that belongs
         # to a different task — pretend it doesn't exist.
         if not info.args or info.args[0] != task_id:
+            logger.warning(
+                "Merge job not found: args mismatch merge_job_id=%s task_id=%s info_args=%r",
+                merge_job_id,
+                task_id,
+                info.args,
+            )
             raise HTTPException(
                 status_code=404, detail=f"Merge job {merge_job_id} not found"
             )
@@ -795,6 +819,11 @@ async def get_merge_job(
         if status_str is None:
             # Race: arq evicted the job's status entry between our info()
             # call and now. Treat as not-found.
+            logger.warning(
+                "Merge job not found: status=None after info OK merge_job_id=%s task_id=%s",
+                merge_job_id,
+                task_id,
+            )
             raise HTTPException(
                 status_code=404, detail=f"Merge job {merge_job_id} not found"
             )

backend/src/workers/job_queue.py

@@ -92,8 +92,16 @@ def _job(cls, pool: ArqRedis, job_id: str) -> Job:
         Job(job_id=..., redis=pool) constructor. The Job handle is
         cheap to create — it's just a pair of references — and reading
         from Redis happens lazily on info()/status()/result().
+
+        We pin `_queue_name` to our default queue so Job.status() can
+        find queued-but-not-yet-started jobs (status() reads the queue
+        ZSET via `zscore(self._queue_name, job_id)` to detect the
+        `queued` state). Without this, status() would default to
+        arq's `arq:queue` and return `not_found` for anything still
+        waiting in our `supoclip_tasks` queue. info() and result() are
+        queue-agnostic.
         """
-        return Job(job_id=job_id, redis=pool)
+        return Job(job_id=job_id, redis=pool, _queue_name=DEFAULT_QUEUE_NAME)
 
     @classmethod
     async def get_job_result(cls, job_id: str):