diff --git a/docker/Dockerfile b/docker/Dockerfile
index d94146ae88824..b59cb924cc795 100644
--- a/docker/Dockerfile
+++ b/docker/Dockerfile
@@ -1,4 +1,4 @@
-FROM ubuntu:22.04
+FROM debian:bookworm
 
 ARG TARGETPLATFORM
 
diff --git a/docker/debian/meta.Dockerfile b/docker/debian/meta.Dockerfile
index ffd030c43cbe4..821c9d16202cf 100644
--- a/docker/debian/meta.Dockerfile
+++ b/docker/debian/meta.Dockerfile
@@ -1,4 +1,5 @@
-FROM debian:bullseye
+FROM debian:bookworm
+
 ARG TARGETPLATFORM
 ENV TERM=dumb
 ENV DEBIAN_FRONTEND=noninteractive
diff --git a/docker/debian/query.Dockerfile b/docker/debian/query.Dockerfile
index dc44202ecad18..28f49f2e22636 100644
--- a/docker/debian/query.Dockerfile
+++ b/docker/debian/query.Dockerfile
@@ -1,4 +1,5 @@
-FROM debian:bullseye
+FROM debian:bookworm
+
 ARG TARGETPLATFORM
 ENV TERM=dumb
 ENV DEBIAN_FRONTEND=noninteractive
diff --git a/docs/doc/13-sql-clients/01-bendsql.md b/docs/doc/13-sql-clients/01-bendsql.md
index 3ecf54f4c0d2c..b2230baac8e08 100644
--- a/docs/doc/13-sql-clients/01-bendsql.md
+++ b/docs/doc/13-sql-clients/01-bendsql.md
@@ -136,8 +136,21 @@ from file:
     --format-opt="skip_header=1" \
     --set="presigned_url_disabled=1" \
     --data=@ontime.csv.gz
+
+> bendsql \
+    --query='REPLACE INTO sample VALUES;' \
+    --format=csv \
+    --format-opt="compression=auto" \
+    --format-opt="skip_header=0" \
+    --format-opt="field_delimiter=," \
+    --format-opt="record_delimiter=\n" \
+    --data="@sample.csv.xz"
 ```
 
 :::note
 `presigned_url_disabled=1` would instruct BendSQL to load data directly using `upload_to_stage` api, which would result in additional transfer fee as well as lower performance, and is not recommended for production use.
-:::
\ No newline at end of file
+:::
+
+:::tip
+For more information about `--format-opt`, please refer to [Input & Output File Formats](../13-sql-reference/50-file-format-options.md).
+:::