Honor zfs_dataset = "none" instead of erroring; surface shutdown-prep panel from fleet view

Two problems surfaced while checking on the fleet after the shutdown-
prep button PR:

1. morpheus is configured with zfs_dataset = "none" (it has no ZFS
   pool -- warm tier is a plain directory on the LVM root volume),
   but nothing in the code actually implemented that as a sentinel.
   cmd_snapshot/cmd_replicate and the daemon's periodic snap/repl
   ticks always tried real zfs/zpool calls regardless, producing
   "zfs: command not found" errors on every hourly tick and in the
   shutdown-prep report. WarmConfig::zfs_enabled() now gates all four
   call sites; a non-ZFS node gets a clean "nothing to
   snapshot/replicate" instead of a raw shell error.

2. safe-shutdown-prep.sh's zpool-health step now checks `command -v
   zpool` first instead of leaking "zpool: command not found" into
   the report.

3. "we don't see the button" turned out to be page confusion: the
   shutdown-prep panel lives on the per-node detail page
   (/v2/nodes/<name>), not the root Fleet Health landing page. Added
   a small "view detail · maintenance & shutdown prep →" hint to the
   bottom of every NodeCard so it's discoverable without already
   knowing to click through.

Verified against tank, architect, and morpheus -- morpheus's
shutdown-prep --dry-run report is now clean (no "command not found"
lines) both when run locally and via cross-node RPC from tank.

Co-Authored-By: Claude Sonnet 5 <[email protected]>

claw-store/src/config.rs

@@ -27,12 +27,26 @@ pub struct HotConfig {
 #[derive(Debug, Clone, Deserialize, Serialize)]
 pub struct WarmConfig {
     pub projects_path: PathBuf,
+    /// Dataset name for `zfs`/`zpool` operations against the warm
+    /// tier. The literal value `"none"` means this node's warm tier
+    /// is a plain directory, not ZFS-backed (e.g. a build node with
+    /// no ZFS pool) — snapshot/replicate become no-ops instead of
+    /// erroring on a missing `zfs`/`zpool` binary. See
+    /// [`WarmConfig::zfs_enabled`].
     pub zfs_dataset: String,
     pub snapshot_retain_hours: u64,
     pub snapshot_retain_days: u64,
     pub snapshot_retain_weeks: u64,
 }
 
+impl WarmConfig {
+    /// `false` when `zfs_dataset = "none"` — this node's warm tier
+    /// has no ZFS pool underneath it.
+    pub fn zfs_enabled(&self) -> bool {
+        self.zfs_dataset != "none"
+    }
+}
+
 #[derive(Debug, Clone, Deserialize, Serialize)]
 pub struct ColdConfig {
     pub archive_path: PathBuf,

claw-store/src/daemon.rs

@@ -188,6 +188,9 @@ pub async fn run(cfg: Config, mut manifest: Manifest) -> Result<()> {
                 }
             }
             _ = snap_tick.tick() => {
+                if !cfg.warm.zfs_enabled() {
+                    tracing::debug!("skipping snapshot tick — zfs_dataset = \"none\" on this node");
+                } else {
                     let ts = chrono::Utc::now().format("%Y-%m-%d-%H%M").to_string();
                     tracing::info!("taking snapshot {}", ts);
                     if let Err(e) = snapshot::run_snapshot_cycle(
@@ -199,8 +202,11 @@ pub async fn run(cfg: Config, mut manifest: Manifest) -> Result<()> {
                         tracing::error!("snapshot failed: {:#}", e);
                     }
                 }
+            }
             _ = repl_tick.tick() => {
-                if let Some(rep) = &cfg.replication {
+                if !cfg.warm.zfs_enabled() {
+                    tracing::debug!("skipping replication tick — zfs_dataset = \"none\" on this node");
+                } else if let Some(rep) = &cfg.replication {
                     if let (Some(host), Some(user), Some(dest)) = (
                         &rep.send_to_host, &rep.send_to_user, &rep.cold_dataset_on_peer
                     ) {

claw-store/src/main.rs

@@ -1502,6 +1502,10 @@ fn cmd_gc(cfg: &Config, manifest: &mut Manifest) -> Result<()> {
 }
 
 fn cmd_snapshot(cfg: &Config, zfs: &SystemZfs) -> Result<()> {
+    if !cfg.warm.zfs_enabled() {
+        println!("Warm tier is not ZFS-backed on this node (zfs_dataset = \"none\") — nothing to snapshot.");
+        return Ok(());
+    }
     let ts = chrono::Utc::now().format("%Y-%m-%d-%H%M").to_string();
     snapshot::run_snapshot_cycle(
         zfs, &cfg.warm.zfs_dataset, &ts,
@@ -1514,6 +1518,10 @@ fn cmd_snapshot(cfg: &Config, zfs: &SystemZfs) -> Result<()> {
 }
 
 fn cmd_list_snapshots(cfg: &Config, zfs: &SystemZfs, _project: &str) -> Result<()> {
+    if !cfg.warm.zfs_enabled() {
+        println!("Warm tier is not ZFS-backed on this node (zfs_dataset = \"none\").");
+        return Ok(());
+    }
     let snaps = zfs.list_snapshots(&cfg.warm.zfs_dataset)?;
     if snaps.is_empty() { println!("No snapshots found."); return Ok(()); }
     for s in &snaps { println!("  {}", s); }
@@ -1531,6 +1539,10 @@ fn cmd_restore(cfg: &Config, zfs: &SystemZfs, project: &str, snap: &str) -> Resu
 }
 
 fn cmd_replicate(cfg: &Config, zfs: &SystemZfs) -> Result<()> {
+    if !cfg.warm.zfs_enabled() {
+        println!("Warm tier is not ZFS-backed on this node (zfs_dataset = \"none\") — nothing to replicate.");
+        return Ok(());
+    }
     let rep = cfg.replication.as_ref()
         .context("no replication config — this node does not replicate")?;
     let host = rep.send_to_host.as_ref().context("send_to_host not set")?;

dashboard-v2/src/components/NodeCard.js

@@ -24,7 +24,7 @@ export function NodeCard({ node }) {
                 border,
             ].join(' '), children: [_jsxs("div", { className: "flex items-center justify-between", children: [_jsxs("div", { className: "flex items-center gap-2", children: [_jsx("span", { className: `inline-block w-2.5 h-2.5 rounded-full ${dot}` }), _jsx("span", { className: "text-lg font-semibold text-slate-100", children: node.node_name })] }), _jsx("span", { className: "text-xs text-slate-500 font-mono", children: node.zone || '—' })] }), !node.online && (_jsx("div", { className: "text-sm text-red-400 break-words", children: node.error ?? 'offline' })), node.online && (_jsxs(_Fragment, { children: [node.filesystem && (_jsx(StorageBar, { label: "disk", used: node.filesystem.used_bytes, total: node.filesystem.total_bytes })), node.hot && node.hot.max_bytes > 0 && (_jsx(StorageBar, { label: "hot tier", used: node.hot.used_bytes, total: node.hot.max_bytes, pinned: node.hot.pinned_bytes ?? undefined })), _jsxs("div", { className: "grid grid-cols-2 gap-y-1 text-sm", children: [_jsx("span", { className: "text-slate-500", children: "mount" }), _jsx("span", { className: "text-right font-mono text-xs", children: node.mount?.active ? (_jsx("span", { className: "text-emerald-300", children: "\u2713 mounted" })) : (_jsx("span", { className: "text-slate-500", children: "not mounted" })) }), _jsx("span", { className: "text-slate-500", children: "cache hit rate" }), _jsx("span", { className: "text-right font-mono text-xs", children: node.cache && node.cache.hits + node.cache.misses > 0
                                         ? `${Math.round(node.cache.hit_rate * 100)}%`
-                                        : _jsx("span", { className: "text-slate-500", children: "idle" }) }), _jsx("span", { className: "text-slate-500", children: "next scheduled job" }), _jsx("span", { className: "text-right font-mono text-xs", children: nextTimer(node) })] })] }))] }) }));
+                                        : _jsx("span", { className: "text-slate-500", children: "idle" }) }), _jsx("span", { className: "text-slate-500", children: "next scheduled job" }), _jsx("span", { className: "text-right font-mono text-xs", children: nextTimer(node) })] })] })), _jsx("div", { className: "pt-1 border-t border-slate-800 text-xs text-slate-500", children: "view detail \u00B7 maintenance & shutdown prep \u2192" })] }) }));
 }
 function healthOf(n) {
     if (!n.online)

dashboard-v2/src/components/NodeCard.tsx

@@ -95,6 +95,10 @@ export function NodeCard({ node }: Props) {
             </div>
           </>
         )}
+
+        <div className="pt-1 border-t border-slate-800 text-xs text-slate-500">
+          view detail · maintenance & shutdown prep →
+        </div>
       </a>
     </Link>
   );

deploy/scripts/safe-shutdown-prep.sh

@@ -211,14 +211,21 @@ step "flushing filesystem buffers"
 sync
 ok "sync complete"
 
-POOL=$(df --output=source /slab 2>/dev/null | tail -1 | tr -d '[:space:]')
+step "zpool health check"
-if [ -n "$POOL" ] && [ "$POOL" != "none" ]; then
+if ! command -v zpool >/dev/null 2>&1; then
-  step "zpool health check ($POOL)"
+  ok "no zpool binary on this node — warm tier is not ZFS-backed here, nothing to check"
+else
+  POOL=$(df --output=source /slab 2>/dev/null | tail -1 | tr -d '[:space:]')
+  if [ -n "$POOL" ] && [ "$POOL" != "none" ]; then
+    echo "  pool: $POOL"
     zpool status -x "$POOL" 2>&1 | sed 's/^/  /'
     if [ "$EXPORT_ZPOOL" -eq 1 ]; then
       step "exporting $POOL (--export-zpool)"
       run sudo zpool export "$POOL" && ok "exported — remember: zpool import $POOL after reboot" || fail "export failed"
     fi
+  else
+    warn "zpool present but /slab isn't a recognizable ZFS mount"
+  fi
 fi
 
 hr