Accounts - Reassign datasets

user031170

I would like some additional functionality built into the reassign option. Specifically, functionality to allow you to pick what datasets you want to update. Currently it is all or none, which can include datasets you don't own, which will fail and give a generic error message. I was attempting to flip 100 datasets to a new one and would have loved to test with a few and only update the ones I own.

jaeW_at_Onyx

i had my EmmaBot (the Domo User Group community chatbot slap together this solution)

i don't know if you're in our slack, but it's a free community. the slack post is here

https://join.slack.com/t/domousergroup/shared_invite/zt-3u8n0h9sh-igBlIJS_TcAR~7sr7xYXmQ

https://domousergroup.slack.com/archives/C0AQRRBUFPB/p1788981360704129?thread_ts=1788979685.511789&cid=C0AQRRBUFPB

The all-or-none reassign is genuinely painful at scale — but the good news is the API gives you exactly what you’re asking for: pick which DataSets to flip, test with a few first, and see which ones would fail (and why) instead of a generic error.

How reassignment actually works

Connector-backed DataSets don’t store the account on the DataSet itself — the account lives on the DataSet’s stream (the connector configuration). The Accounts UI “reassign” updates that stream config for every DataSet on the account, which is why one DataSet you can’t edit kills the whole batch:

1. GET /api/data/v3/datasources/{dataset_id} — DataSet metadata: streamId, transportType, owner, and currentUserFullAccess
2. GET /api/data/v1/streams/{stream_id} — full connector config, including the account object
3. PUT /api/data/v1/streams/{stream_id} — write the config back with the new account.id

Reassigning one DataSet = one PUT. Loop over only the ones you want.

Script (dry-run by default)

#!/usr/bin/env python3
"""
Reassign Domo connector DataSets to a different account — selectively.
Dry-run by default: nothing is modified unless you pass --execute.

Setup:
  1. Create a client at developer.domo.com with scopes: data, account
  2. Exchange it for an access token (client credentials grant)
  3. Grab the numeric account ID from the account's page in the Accounts UI
     (it's in the URL), and your list of DataSet IDs (the account's DataSets
     list, Data Center, or your own inventory)
"""
import argparse
from pathlib import Path

import requests

PUT_FIELDS = ("transport", "configuration", "updateMethod", "dataProvider", "dataSource")


def get_dataset(base, headers, dataset_id):
    r = requests.get(f"{base}/api/data/v3/datasources/{dataset_id}", headers=headers)
    r.raise_for_status()
    return r.json()


def get_stream(base, headers, stream_id):
    r = requests.get(f"{base}/api/data/v1/streams/{stream_id}", headers=headers)
    r.raise_for_status()
    return r.json()


def reassign_stream(base, headers, stream, new_account_id):
    body = {field: stream[field] for field in PUT_FIELDS}
    body["account"] = {"id": new_account_id}
    if stream.get("advancedScheduleJson"):
        body["advancedScheduleJson"] = stream["advancedScheduleJson"]
    return requests.put(f"{base}/api/data/v1/streams/{stream['id']}",
                        headers=headers, json=body)


def main():
    p = argparse.ArgumentParser()
    p.add_argument("--instance", required=True, help="e.g. 'mycompany' for mycompany.domo.com")
    p.add_argument("--token", required=True, help="access token (data + account scopes)")
    p.add_argument("--new-account", type=int, required=True)
    p.add_argument("--ids", required=True,
                   help="comma-separated DataSet IDs, or a path to a file with one per line")
    p.add_argument("--execute", action="store_true", help="actually reassign (default: dry-run)")
    args = p.parse_args()

    base = f"https://{args.instance}.domo.com"
    headers = {"X-DOMO-Developer-Token": args.token, "Content-Type": "application/json"}

    if Path(args.ids).is_file():
        ids = [line.strip() for line in Path(args.ids).read_text().splitlines() if line.strip()]
    else:
        ids = [i.strip() for i in args.ids.split(",")]

    would, done, skipped, failed = 0, 0, 0, 0
    for ds_id in ids:
        try:
            ds = get_dataset(base, headers, ds_id)
        except requests.HTTPError as e:
            print(f"✗ {ds_id}: not readable ({e.response.status_code})")
            failed += 1
            continue

        name, owner = ds.get("name", "?"), ds.get("owner", {}).get("name", "?")

        if ds.get("transportType") != "CONNECTOR" or not ds.get("streamId"):
            print(f"- {ds_id} '{name}': skipped — not a connector DataSet (no account)")
            skipped += 1
            continue

        stream = get_stream(base, headers, ds["streamId"])
        old_acct = (stream.get("account") or {}).get("id")

        if not ds.get("currentUserFullAccess"):
            # this is the DataSet that breaks the UI's all-or-none reassign
            print(f"! {ds_id} '{name}' (owner: {owner}): NO EDIT ACCESS — would fail, skipping")
            skipped += 1
            continue

        if not args.execute:
            print(f"~ {ds_id} '{name}': would reassign account {old_acct} → {args.new_account}")
            would += 1
            continue

        r = reassign_stream(base, headers, stream, args.new_account)
        if r.ok:
            print(f"✓ {ds_id} '{name}': reassigned {old_acct} → {args.new_account}")
            done += 1
        else:
            print(f"✗ {ds_id} '{name}': {r.status_code} {r.text[:150]}")
            failed += 1

    mode = "EXECUTED" if args.execute else "DRY RUN — pass --execute to apply"
    print(f"\n{mode}: {would or done} to change, {skipped} skipped, {failed} failed")


if __name__ == "__main__":
    main()

Workflow

# 1. preview the whole list — see what would change and what you can't edit
python reassign_datasets.py --instance mycompany --token $TOKEN \
    --new-account 456 --ids all_100.txt

# 2. test with a few for real
python reassign_datasets.py --instance mycompany --token $TOKEN \
    --new-account 456 --ids "ds-aaa,ds-bbb,ds-ccc" --execute

# 3. flip the rest
python reassign_datasets.py --instance mycompany --token $TOKEN \
    --new-account 456 --ids all_100.txt --execute

Caveats

• The new account must be for the same connector — you can’t point a Google Sheets DataSet at a Salesforce account.
• You need Edit DataSet rights on each DataSet — the dry-run flags the ones where you don’t (currentUserFullAccess: false). Those are exactly the ones that make the UI reassign die with a generic error.
• Magic ETL outputs, file uploads, and webforms have no account — the script skips them automatically.
• PUT /api/data/v1/streams/{id} is documented but less forgiving than the public DataSet API — if your instance rejects the payload, trim dataSource down to {id, name, description}.

Docs:

• Datasets API (includes Stream Get/Update): https://www.domo.com/docs/portal/35115c5c48927-datasets-api
• Manage Connector Accounts (the UI you’re using today): https://www.domo.com/docs/s/article/360042926054