pageserver: don't try and recover deletion queue if no remote storage (#5419)

## Problem

Because `neon_local` by default runs with no remote storage, it was not
running the deletion queue workers, and the attempt to call into
`recover()` was failing.

This is a bogus configuration that will go away when we make remote
storage mandatory.

## Summary of changes

Don't try and do deletion queue recovery when remote storage is
disabled.

The reason we don't just unset `control_plane_api` to avoid this is that
generations will soon become mandatory, irrespective of when we make
remote storage mandatory.

pageserver/src/tenant/mgr.rs

@@ -130,10 +130,15 @@ pub async fn init_tenant_mgr(
         // deletion list entries may still be valid.  We provide that by pushing a recovery operation into
         // the queue. Sequential processing of te queue ensures that recovery is done before any new tenant deletions
         // are processed, even though we don't block on recovery completing here.
-        resources
-            .deletion_queue_client
-            .recover(result.clone())
-            .await?;
+        //
+        // Must only do this if remote storage is enabled, otherwise deletion queue
+        // is not running and channel push will fail.
+        if resources.remote_storage.is_some() {
+            resources
+                .deletion_queue_client
+                .recover(result.clone())
+                .await?;
+        }
 
         Some(result)
     } else {