Hello,
My goal is simple: I want to duplicate an existing workbook on
Tableau Cloud and update the custom SQL inside its embedded
datasource — all programmatically via Python.
I am using the Tableau Server Client (TSC) Python library to programmatically
update the custom SQL inside an embedded datasource in a workbook, then
republish it to Tableau Cloud as a new workbook.
**Environment:**
- Tableau Cloud (us-east-1)
- Python tableauserverclient (TSC) library
- Datasource: Amazon Redshift (embedded in workbook, not a published datasource)
- Connection goes through Tableau Bridge (private network/VPN)
**What I am doing:**
1. Download the workbook WITHOUT extract using include_extract=False → get small .twb file
2. Parse the .twb XML and replace the custom SQL in the embedded datasource
3. Republish as a new workbook using server.workbooks.publish()
**The Problem:**
Every publish attempt fails with:
403132: Forbidden
<workbook_name> failed to establish a connection to your datasource.
**Everything I have tried:**
1. Passing ConnectionItem with credentials:
conn_item = ConnectionItem()
conn_item.server_address = "<redshift-host>"
conn_item.server_port = "5439"
conn_item.username = "<username>"
conn_item.password = "<password>"
conn_item.embed_password = True
server.workbooks.publish(..., connections=[conn_item])
2. Using skip_connection_check=True:
server.workbooks.publish(..., skip_connection_check=True)
→ Confirmed skipConnectionCheck=true IS in the POST URL
3. Both combined — still 403132
4. Neither (no connections parameter) — still 403132
5. Confirmed the XML being sent is correct and contains
<connectionCredentials> with the right server address,
username, and embed="true"
**Key observation:**
The original workbook works fine on Tableau Cloud and its extract
refresh runs through Tableau Bridge. Tableau Cloud has no direct
network path to Redshift — it always goes through Bridge.
My hypothesis is that Tableau Cloud ignores skipConnectionCheck
for Bridge-dependent connections and always tries a direct
connection validation, which fails because Redshift is on a
private network.
**Questions:**
1. Is it possible to publish a .twb via REST API when the datasource
connects through Tableau Bridge?
2. Does skipConnectionCheck=true not apply to Bridge connections?
3. Is there a correct way to publish and tell Tableau Cloud to use
the existing Bridge-linked credentials rather than validating
directly?
Code:
server = TSC.Server(SERVER_URL, use_server_version=True)
with server.auth.sign_in(auth):
source_wb = server.workbooks.get_by_id(SOURCE_WORKBOOK_ID)
buf = io.BytesIO()
server.workbooks.download(
SOURCE_WORKBOOK_ID,
filepath=buf,
include_extract=False,
)
buf.seek(0)
new_wb_item = TSC.WorkbookItem(
name=NEW_WORKBOOK_NAME,
project_id=TARGET_PROJECT_ID,
)
new_wb_item.show_tabs = source_wb.show_tabs
#
# # 5. Publish using the modified source connection
new_wb = server.workbooks.publish(
new_wb_item,
buf,
mode=TSC.Server.PublishMode.CreateNew,
# connections=[source_conn], # <-- Tried with conn also
skip_connection_check=True
)
Thank you.
Hi @DIEGO FERNANDO MARTINEZ RODRIGUEZ I don't want to go with `extract` files as we have to migrate 100+ workbook and each workbook witch extrat size is arround 2GB+. So if there is other way to replace query without download then happy to use that. Else I am okay to donlowad but without extact as file size is huge.
