diff --git a/product_docs/docs/pgd/4/images/bdr-ao-spec.png b/product_docs/docs/pgd/4/images/bdr-ao-spec.png
new file mode 100644
index 00000000000..45d87bc5c20
--- /dev/null
+++ b/product_docs/docs/pgd/4/images/bdr-ao-spec.png
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:68f18458d6e1bc04715e6a64381c39e142b9d09ac7321eba0bf2d79c2e4ba5d2
+size 44744
diff --git a/product_docs/docs/pgd/4/overview/index.mdx b/product_docs/docs/pgd/4/overview/index.mdx
index e0c678ee8c5..fdf74eb0640 100644
--- a/product_docs/docs/pgd/4/overview/index.mdx
+++ b/product_docs/docs/pgd/4/overview/index.mdx
@@ -449,7 +449,7 @@ nodes. This does not count any Logical Standby nodes.
 The current and standard representation of this can be seen in the following
 diagram:
 
-![BDR Always-On Reference Architecture](images/bdr-ao-spec.png)
+![BDR Always-On Reference Architecture](../images/bdr-ao-spec.png)
 
 In this diagram, HARP Manager would exist on BDR Nodes 1-4. The initial state
 of the cluster would be that BDR Node 1 is the Lead master of DC A, and BDR