feat(translation): pooled Interpreter session + manual sweep resumes after drain

- interpreter_client: shared requests.Session with a connect-only retry
  (connect=2, no status retries — we map 429/5xx ourselves) so a proxy reload
  is smoothed and the keep-alive connection is pooled across the sweep.
- translate_posts: on an interrupt (drain), re-enqueue after the Retry-After
  hint / default backoff instead of waiting for the daily beat; self-terminates
  via the health gate. Steady-state one-chunk-per-run on success is unchanged.

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>

tests/test_translate_posts.py

@@ -137,6 +137,31 @@ def test_translate_posts_service_down_leaves_untranslated(db_sync, monkeypatch):
     assert p.translated_source_lang is None       # will retry next run
 
 
+def test_translate_posts_interrupt_reenqueues_after_backoff(db_sync, monkeypatch):
+    # A drain mid-sweep (health passed, translate 503s w/ Retry-After) re-enqueues
+    # the daily sweep after the backoff instead of waiting for tomorrow's beat.
+    _patch(monkeypatch, db_sync)
+    monkeypatch.setattr("backend.app.tasks.translation.ic.health", lambda *a, **k: True)
+
+    def _drain(texts, **k):
+        raise ic.InterpreterUnavailable("draining", retry_after=30)
+
+    monkeypatch.setattr("backend.app.tasks.translation.ic.translate", _drain)
+    calls = []
+    monkeypatch.setattr(
+        "backend.app.tasks.translation.translate_posts.apply_async",
+        lambda *a, **k: calls.append((a, k)),
+    )
+    a = _artist(db_sync)
+    _post(db_sync, a.id, title="ねこ", ext="p1")
+    _enable(db_sync)
+    db_sync.commit()
+
+    assert "interrupted" in translate_posts()
+    assert len(calls) == 1
+    assert calls[0][1]["countdown"] == 30
+
+
 def _mock_new_model(monkeypatch, *, ver="v2"):
     """Interpreter is up and translates with a NEW engine version."""
     monkeypatch.setattr("backend.app.tasks.translation.ic.health", lambda *a, **k: True)