nodepool: revert dedicated diskimages list for rax-iad

When we had some issues with image uploads to rax-iad last year, we
needed to configure it with a dedicated list of diskimages in order to
be able to add "pause: true" statements to those. When the issue was
resolved, we only dropped the pause statements instead of reverting to
use the original list shared across all providers [0]. This is getting
fixed now.

[0] Ia2b737e197483f9080b719bab0ca23461850e157

Change-Id: Ibc1a11d04d5258ca503f8cd5b5fe7a30eb99bc56

nodepool/nodepool.yaml

@@ -68,33 +68,7 @@ providers:
     rate: 0.001
     # 6 hours
     image-upload-timeout: 21600
-    diskimages:
-      - name: centos-8-stream
-        config-drive: true
-      - name: centos-9-stream
-        config-drive: true
-      - name: debian-bookworm
-        config-drive: true
-      - name: debian-bullseye
-        config-drive: true
-      - name: gentoo-17-0-systemd
-        config-drive: true
-      - name: rockylinux-8
-        config-drive: true
-      - name: rockylinux-9
-        config-drive: true
-      - name: ubuntu-bionic
-        config-drive: true
-      - name: ubuntu-focal
-        config-drive: true
-      - name: ubuntu-jammy
-        config-drive: true
-      - name: ubuntu-noble
-        config-drive: true
-      - name: ubuntu-xenial
-        config-drive: true
-      - name: openEuler-22-03-LTS
-        config-drive: true
+    diskimages: *provider_diskimages
 
   - name: rax-ord
     region-name: 'ORD'