hi all, I'm having an error while ingesting MSSQL data into the datahub. It says db connection url couldn't be parsed. Below is the error returned. Also I shared my ingestion yaml file here. Am I missing something?

source:

type: mssql

config:

# Coordinates

host_port: '192.168.1.1:1433'

database: 'prod-db-2'

scheme: 'PROD-DB2'

# Credentials

username: db2admin

password: "password*"

sink:

type: "datahub-rest"

config:

server: "<http://localhost:6080>"

transformers:

- type: "simple_add_dataset_tags"

config:

tag_urns:

- "urn:li:tag:db2"

ArgumentError: Could not parse rfc1738 URL from string '<PROD-DB2://db2admin:password%2A@192.168.1.1:1433/prod-db-2>'
[2022-07-06 01:26:39,723] INFO     {datahub.entrypoints:176} - DataHub CLI version: 0.8.34.2 at /usr/local/lib/python3.8/dist-packages/datahub/__init__.py
[2022-07-06 01:26:39,723] INFO     {datahub.entrypoints:179} - Python version: 3.8.10 (default, Mar 15 2022, 12:22:08)
[GCC 9.4.0] at /usr/bin/python3 on Linux-5.4.0-104-generic-x86_64-with-glibc2.29
[2022-07-06 01:26:39,723] INFO     {datahub.entrypoints:182} - GMS config {'models': {}, 'versions': {'linkedin/datahub': {'version': 'v0.8.34', 'commit': 'f847fa31c9010bbb9df0d13ae7660e59083ea03e'}}, 'managedIngestion': {'defaultCliVersion': '0.8.34.1', 'enabled': True}, 'statefulIngestionCapable': True, 'supportsImpactAnalysis': True, 'telemetry': {'enabledCli': True, 'enabledIngestion': False}, 'datasetUrnNameCasing': False, 'retention': 'true', 'noCode': 'true'}

I believe, you will need to download and install ODBC driver for SQL Server first. Follow instructions for your OS - https://docs.microsoft.com/en-gb/sql/connect/odbc/download-odbc-driver-for-sql-server?view=sql-server-ver16

Yes, already installed and I got different error message.

Hey @gray-architect-29447 what was the error ? Were you able to run ingestion successfully ?

Hi @hundreds-photographer-13496, I just found that MSSQL server authentication mode was set to Windows Integrated Authentication. So I've changed it to the mixed mode. Now I'm getting

"could not find a login matching the name provided sql connection"

error. Since, the user is Windows domain user and has @company.local prefix, I need to change username to : "COMPANYLOCAL\dbadmin". But it will also result an error.

can you try escaping the backslash like this "COMPANYLOCAL\\dbadmin"

It says login failed for user COMPANYLOCAL\\dbadmin

Seems we are past the earlier error of "`could not find a login matching the name provided sql connection`"

Oh ! Looks like there an open github issue for this - https://github.com/datahub-project/datahub/issues/5344

nice, gonna follow this issue

Cool, meanwhile, you could also try ingestion without using odbc, by removing

use_odbc

and

uri_args

and check if it works. (you will also need to remove

scheme

if you haven't already)

woohoo, followed your guide and data ingested successfully. But I don't get how it's done without odbc, at the beginning it was asking for to use odbc. Now it's working without odbc

Yay!! did it ask for odbc explicitly ?? I thought you had to switch to it due to url parsing error

It didn't ask for odbc explicitly. Everything went great without any error. Thanks 🙂

great !

@hundreds-photographer-13496 Hi Mayuri, I am stuck with a similar error. I couldn't get past the login for windows authentication. (Sql auth works fine ) Unfortunately, I cannot use sql logins.

Hi @flaky-soccer-57765 are you getting same error as this github issue ? Looks like its still open.

@hundreds-photographer-13496 No, I am looking for general guidance around making windows authentication work. As of now, only SQL server login works. Attaching the error log and recipe used.

Hi @flaky-soccer-57765 looks like you are executing the recipe on managed ingestion. Can you confirm - if its running, when using datahub cli alone?

@hundreds-photographer-13496 I haven't tried that. Will try and come back.

Hi @flaky-soccer-57765 - the second error kerberos not found is for managed ingestion. I would suggest - to first successfully execute recipe with cli ingestion on local and then attempt managed ingestion, because managed ingestion runs on a separate container, which is an added complexity to get it to work.