scheduler: Shorten tolerations for node failure

Similar to what was done in #1055, we need to explicitly add tolerations
to the scheduler to get it to be recreated more quickly on node failure.

This is particularly necessary because we don't have #955. We could wait
for that, but it's a lot of work, and this is a small thing we can do in
the meantime.

Fixes neondatabase/cloud#17298.

autoscale-scheduler/deployment.yaml

@@ -63,3 +63,13 @@ spec:
         - name: plugin-config-volume
           configMap:
             name: scheduler-plugin-config
+
+      tolerations:
+        # Add explicit (short) tolerations for node failure, because otherwise the default of 5m
+        # will be used, which is unacceptably long for us.
+        - key: node.kubernetes.io/not-ready
+          tolerationSeconds: 30
+          effect: NoExecute
+        - key: node.kubernetes.io/unreachable
+          tolerationSeconds: 30
+          effect: NoExecute