feat(docker): swarm-aware container views — collapse replicas, show host, surface agent-less tasks

The container list and dashboard widget listed each swarm task as its own cryptic
row (svc.1.<taskid>) and could only show tasks on nodes that run a Steward agent.
Make both views service-centric and cluster-complete.

- swarm_view.build_swarm_services(): model-free builder that merges real container
  rows (collected local-per-node, so host_id = the node a task runs on) with the
  managers' placement (DockerSwarmService.placement_json). Where placement counts
  exceed the local rows on a node — i.e. nodes with no agent — it synthesizes
  "ghost" replicas so the replica list matches the cluster. Non-swarm containers
  pass through grouped by host.
- /rows + rows.html: a Swarm-services panel collapses each service to one block,
  every replica a host chip (status + cpu); ghosts render dashed "N · no agent".
  Non-swarm/compose containers keep the per-host table below. New Services stat.
- /widget + widget.html: same service collapse with host chips; standalone
  containers stay grouped by host.
- Unit tests for the builder (real+ghost merge, full coverage, partial agent-less,
  synthesized service, down state).

No migration — node_id, placement_json and swarm-node hostnames are already stored.

Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_016Jg27rgypiW2efULXJDtMC

plugins/docker/routes.py

@@ -11,6 +11,7 @@ from steward.models.users import UserRole
 from steward.models.hosts import Host
 from steward.core.time_range import parse_range, DEFAULT_RANGE
 from .dedup import dedup_by_container_id
+from .swarm_view import build_swarm_services
 from .models import (
     DockerContainer, DockerDiskUsage, DockerEvent, DockerImage, DockerMetric,
     DockerSwarmNode, DockerSwarmService,
@@ -138,7 +139,8 @@ async def _host_map(db, host_ids: set[str]) -> dict[str, Host]:
 @docker_bp.get("/rows")
 @require_role(UserRole.viewer)
 async def rows():
-    """HTMX fragment: containers grouped by host, with resource sparklines."""
+    """HTMX fragment: swarm services (collapsed, cluster-complete) on top, then
+    non-swarm containers grouped by host with resource sparklines."""
     since, range_key = parse_range(request.args.get("range"))
 
     async with current_app.db_sessionmaker() as db:
@@ -148,12 +150,24 @@ async def rows():
                 DockerContainer.name,
             )
         )).scalars())
-        hosts = await _host_map(db, {c.host_id for c in containers})
+        services = list((await db.execute(select(DockerSwarmService))).scalars())
+        nodes = list((await db.execute(select(DockerSwarmNode))).scalars())
+        hosts = await _host_map(
+            db,
+            {c.host_id for c in containers}
+            | {n.host_id for n in nodes} | {s.host_id for s in services},
+        )
 
-        # Group by host so each container is clearly attributed to the box it
-        # runs on (names are only unique within a host now).
+        # Swarm tasks collapse into per-service rows (each replica labelled with
+        # its host, plus "ghost" replicas for tasks on agent-less nodes, drawn
+        # from the managers' placement). Non-swarm containers keep the host view.
+        node_hostname = {n.node_id: n.hostname for n in nodes if n.hostname}
+        host_name = {h.id: h.name for h in hosts.values()}
+        swarm_view = build_swarm_services(containers, services, node_hostname, host_name)
+
+        standalone = [c for c in containers if not c.service_name]
         groups: dict[str, dict] = {}
-        for c in containers:
+        for c in standalone:
             g = groups.get(c.host_id)
             if g is None:
                 host = hosts.get(c.host_id)
@@ -176,15 +190,15 @@ async def rows():
             else:
                 g["stopped"] += 1
 
-    # Sub-group each host's containers by compose project (or swarm service),
-    # preserving the running-first ordering. Containers with neither label fall
-    # into an unlabelled bucket rendered flat, so plain hosts look unchanged.
+    # Sub-group each host's containers by compose project, preserving the
+    # running-first ordering. Containers with no project fall into an unlabelled
+    # bucket rendered flat, so plain hosts look unchanged.
     for g in groups.values():
         subs: dict[str, dict] = {}
         order: list[str] = []
         for item in g["containers"]:
             c = item["container"]
-            label = c.compose_project or c.service_name or None
+            label = c.compose_project or None
             key = label or ""
             if key not in subs:
                 subs[key] = {"label": label, "containers": []}
@@ -194,14 +208,16 @@ async def rows():
         g["grouped"] = any(s["label"] for s in g["subgroups"])
 
     host_groups = sorted(groups.values(), key=lambda g: g["host_name"].lower())
-    running = sum(g["running"] for g in host_groups)
-    total = len(containers)
+    swarm_services = swarm_view["services"]
+    standalone_running = sum(g["running"] for g in host_groups)
+    ghost_total = sum(r["count"] for s in swarm_services for r in s["replicas"] if r["ghost"])
     return await render_template(
         "docker/rows.html",
         host_groups=host_groups,
-        running=running,
-        stopped=total - running,
-        total=total,
+        swarm_services=swarm_services,
+        running=standalone_running + sum(s["running"] for s in swarm_services),
+        stopped=len(standalone) - standalone_running,
+        total=len(containers) + ghost_total,
         range_key=range_key,
     )
 
@@ -221,6 +237,8 @@ async def widget():
                 DockerContainer.name,
             )
         )).scalars()))
+        services = list((await db.execute(select(DockerSwarmService))).scalars())
+        nodes = list((await db.execute(select(DockerSwarmNode))).scalars())
         # Recent failures are more actionable than a static "stopped" tally. Count
         # DISTINCT container names so the two managers' duplicate die/oom events
         # don't double it.
@@ -228,17 +246,28 @@ async def widget():
             select(func.count(func.distinct(DockerEvent.container_name)))
             .where(DockerEvent.event.in_(("die", "oom")), DockerEvent.at >= since_24h)
         )).scalar() or 0
-        hosts = await _host_map(db, {c.host_id for c in all_containers})
+        hosts = await _host_map(
+            db,
+            {c.host_id for c in all_containers}
+            | {n.host_id for n in nodes} | {s.host_id for s in services},
+        )
 
-    running = [c for c in all_containers if c.status == "running"]
-    display = all_containers if show_stopped else running
+    # Swarm tasks collapse into per-service rows (each replica tagged with its
+    # host); non-swarm containers stay grouped by host.
+    node_hostname = {n.node_id: n.hostname for n in nodes if n.hostname}
+    host_name = {h.id: h.name for h in hosts.values()}
+    swarm_services = build_swarm_services(all_containers, services, node_hostname, host_name)["services"]
+
+    standalone = [c for c in all_containers if not c.service_name]
+    running = [c for c in standalone if c.status == "running"]
+    display = standalone if show_stopped else running
 
     # Group for display so multi-host fleets read clearly; single-host stays flat.
     groups: dict[str, dict] = {}
     for c in display:
         g = groups.setdefault(c.host_id, {
             "host_id": c.host_id,
-            "host_name": hosts[c.host_id].name if c.host_id in hosts else c.host_id,
+            "host_name": host_name.get(c.host_id, c.host_id),
             "containers": [],
         })
         g["containers"].append(c)
@@ -248,9 +277,10 @@ async def widget():
         "docker/widget.html",
         host_groups=host_groups,
         multi_host=len(host_groups) > 1,
-        running_count=len(running),
+        swarm_services=swarm_services,
+        running_count=len(running) + sum(s["running"] for s in swarm_services),
         failed_24h=failed_24h,
-        total_count=len(all_containers),
+        total_count=len(all_containers) + len(swarm_services),
         show_stopped=show_stopped,
         widget_id=widget_id,
     )