Run the pod as non-root with a read-only filesystem

Part of the same security-hardening pass as the last three commits.
Neither the Dockerfile nor the chart's Deployment set any securityContext
at all, so the container ran as root by default — scratch has no
/etc/passwd for a USER directive to resolve against, so nobody had set one.

Dockerfile now ends with USER 65532:65532 (numeric, since scratch has no
user database; 65532 is the common "nonroot" convention, distroless's own
uid). The chart's Deployment adds a matching pod-level securityContext
(runAsNonRoot, runAsUser/runAsGroup: 65532, seccompProfile: RuntimeDefault)
plus per-container hardening (allowPrivilegeEscalation: false,
capabilities dropped, readOnlyRootFilesystem: true) on both the app
container and the wait-for-postgres init container — neither writes
anything to disk, so the root filesystem can stay read-only.

Verified with helm-lint and a manual `helm template` render of both the
terdut-server and terdut-demo charts. Not yet verified: an actual pod
starting with these in place — readOnlyRootFilesystem is exactly where a
non-obvious write (a temp file, a cache dir) would surface as a crash
rather than a lint error, so that needs a real rollout to confirm.

🤖 Generated with [Claude Code](https://claude.com/claude-code)

Co-Authored-By: Claude <noreply@anthropic.com>

Dockerfile

@@ -24,4 +24,10 @@ FROM scratch
 COPY --from=builder /etc/ssl/certs/ca-certificates.crt /etc/ssl/certs/ca-certificates.crt
 COPY --from=builder /terdut /terdut
 EXPOSE 8080
+# Numeric, not a name: scratch has no /etc/passwd for one to resolve against,
+# and Docker's USER accepts a bare UID:GID without it. 65532 is the common
+# "nonroot" convention (distroless's own uid), chosen so the chart's pod
+# securityContext (runAsNonRoot, runAsUser: 65532) matches what the image
+# already runs as rather than fighting it.
+USER 65532:65532
 ENTRYPOINT ["/terdut"]

charts/terdut-server/templates/deployment.yaml

@@ -25,11 +25,30 @@ spec:
         {{- include "terdut-server.selectorLabels" . | nindent 8 }}
     spec:
       enableServiceLinks: false
+      # Pod-wide default; both containers below run as this UID regardless of
+      # what their own image would otherwise pick (postgres:17-alpine's
+      # pg_isready needs no particular user, and 65532 is what the app image
+      # itself runs as now — see the Dockerfile's USER). seccompProfile here
+      # rather than per-container: there is no reason it would ever differ
+      # between them.
+      securityContext:
+        runAsNonRoot: true
+        runAsUser: 65532
+        runAsGroup: 65532
+        seccompProfile:
+          type: RuntimeDefault
       {{- if .Values.database.waitForPostgres.enabled }}
       initContainers:
         - name: wait-for-postgres
           image: "{{ .Values.database.waitForPostgres.image.repository }}:{{ .Values.database.waitForPostgres.image.tag }}"
           imagePullPolicy: {{ .Values.database.waitForPostgres.image.pullPolicy }}
+          # No capability this loop needs, and nothing in it writes to disk:
+          # sh, pg_isready, echo and sleep all run read-only.
+          securityContext:
+            allowPrivilegeEscalation: false
+            readOnlyRootFilesystem: true
+            capabilities:
+              drop: ["ALL"]
           env:
             - name: TERDUT_DB_DSN
               value: {{ required "database.dsn is required" .Values.database.dsn | quote }}
@@ -46,6 +65,13 @@ spec:
         - name: terdut-server
           image: "{{ .Values.image.repository }}:{{ .Values.image.tag }}"
           imagePullPolicy: {{ .Values.image.pullPolicy }}
+          # scratch, nothing to write: the binary keeps no local state and
+          # writes nothing to disk, so the root filesystem can stay read-only.
+          securityContext:
+            allowPrivilegeEscalation: false
+            readOnlyRootFilesystem: true
+            capabilities:
+              drop: ["ALL"]
           ports:
             - name: http
               containerPort: {{ .Values.service.port }}