Merge pull request #376 from OpenMS/claude/demo-workspace-storage-k8s-azhkG

Support configurable demo workspace source directories

Author: t0mdavid-m

clean-up-workspaces.py

@@ -23,6 +23,9 @@
 remaining_directories = []
 # Iterate through directories in workspaces_directory
 for directory in workspaces_directory.iterdir():
+    # Skip hidden top-level directories (e.g. .demos/ holds seeded demo workspaces)
+    if directory.name.startswith("."):
+        continue
     # Check if it's a directory
     if directory.is_dir():
         # Get the directory's modification time

docs/kubernetes-deployment.md

@@ -107,6 +107,21 @@ PersistentVolumeClaim `workspaces-pvc`:
 - `storageClassName: cinder-csi`
 - `resources.requests.storage: 500Gi`
 
+Demo workspaces live under a hidden `.demos/` subdirectory of this PVC (see [Demo workspaces](#demo-workspaces) below). User workspaces live at the PVC root, one directory per session UUID.
+
+### Demo workspaces
+Demo workspaces are seeded onto the `workspaces-pvc` at `/workspaces-streamlit-template/.demos/` by the `seed-demos` initContainer on the Streamlit Deployment. The init runs `cp -rn /app/example-data/workspaces/. /workspaces-streamlit-template/.demos/` — new demos shipped in an image appear after redeploy, but existing entries on the PV (including admin-saved demos and edits) are preserved.
+
+The ConfigMap override points `demo_workspaces.source_dirs` at `/workspaces-streamlit-template/.demos`, so both Streamlit pods and RQ workers read demos from the PV. The "Save as Demo" admin flow writes to the same path.
+
+To force a re-seed of a specific demo, delete it on the PV and restart the Streamlit Deployment:
+```
+kubectl exec deploy/streamlit -- rm -rf /workspaces-streamlit-template/.demos/<name>
+kubectl rollout restart deploy/streamlit
+```
+
+`clean-up-workspaces.py` skips any top-level directory whose name starts with `.`, so the nightly cleanup cron does not touch `.demos/`.
+
 ### `streamlit-deployment.yaml`
 Main Streamlit Deployment. Key fields:
 - `replicas: 2` (scales to N)
@@ -116,6 +131,7 @@ Main Streamlit Deployment. Key fields:
 - Mounts `settings-overrides.json` from the ConfigMap as a `subPath`
 - Readiness and liveness probes hit `/_stcore/health`
 - Pod affinity: `volume-group: workspaces`
+- `seed-demos` initContainer merges image-shipped demos into `.demos/` on the PVC (see [Demo workspaces](#demo-workspaces))
 
 ### `streamlit-service.yaml`
 ClusterIP Service exposing Streamlit on port 8501.

k8s/base/configmap.yaml

@@ -5,5 +5,9 @@ metadata:
 data:
   settings-overrides.json: |
     {
-      "online_deployment": true
+      "online_deployment": true,
+      "demo_workspaces": {
+        "enabled": true,
+        "source_dirs": ["/workspaces-streamlit-template/.demos"]
+      }
     }

k8s/base/streamlit-deployment.yaml

@@ -25,6 +25,19 @@ spec:
                     values:
                       - workspaces
               topologyKey: kubernetes.io/hostname
+      initContainers:
+        - name: seed-demos
+          image: openms-streamlit
+          imagePullPolicy: Always
+          command: ["/bin/sh", "-c"]
+          args:
+            - |
+              set -eu
+              mkdir -p /workspaces-streamlit-template/.demos
+              cp -rn /app/example-data/workspaces/. /workspaces-streamlit-template/.demos/
+          volumeMounts:
+            - name: workspaces
+              mountPath: /workspaces-streamlit-template
       containers:
         - name: streamlit
           image: openms-streamlit

src/common/admin.py

@@ -51,10 +51,25 @@ def get_demo_target_dir() -> Path:
     """
     Get the directory where demo workspaces are stored.
 
+    Reads from settings.demo_workspaces.source_dirs (first entry wins),
+    supporting both the list form and the legacy string form. Falls back
+    to example-data/workspaces when settings are unavailable (e.g. in tests).
+
     Returns:
         Path: The demo workspaces directory.
     """
-    return Path("example-data/workspaces")
+    default = Path("example-data/workspaces")
+    try:
+        demo_config = st.session_state.settings.get("demo_workspaces", {})
+    except (AttributeError, KeyError):
+        return default
+
+    dirs = demo_config.get("source_dirs", demo_config.get("source_dir"))
+    if isinstance(dirs, str):
+        return Path(dirs)
+    if isinstance(dirs, list) and dirs:
+        return Path(dirs[0])
+    return default
 
 
 def demo_exists(demo_name: str) -> bool: