Workbench | Multiple Accounts in One DOMO Instance

cedvill00

Hi,

Is it possible to create multiple accounts to use in Workbench in one instance?

This is because we have different permission/rights based on our roles.

For example, scenario: We are using one server to upload on premise data - for security and refresh schedule.

User 1 has workbench rights and can upload sales data.

User 2 has workbench rights and can only upload hr data but cannot be viewed by user 1.

What is the best way to achieve this, without compromising data security.

GrantSmith

Because how workbench is designed it reads from the same local database so anyone with access to the server can see the different jobs. If you want to keep the two users separate you’d need two separate sever instances for workbench

HowDoIDomo

At my work, we use a local shared workbench server that's always running in the background for reports that need to updated regularly, usually around midnight (since we're rarely logged in at the time). While this specific server uses the same login, we use the "impersonation" function under "configuration" to access the specific file needed for the job despite each user having different permissions.

So, the domo log-in would be the same. But, each person would impersonate themselves for their respective access.

Would this solution work for you?

Edited: Once the impersonation credentials are saved, it's hidden. So, users wouldn't be able to see each others credentials or data if that's one of your concerns.

Jessica

Can I ask how (or can you) you identify which server the Workbench data files coming into your DOMO instance are from?

We are considering loading Workbench on a DEV server, assuming it will still load to our one instance, we want to make sure we can distinguish DEV files from PROD files.

GrantSmith

@Jessica

When I create a dataset I'll make sure to add some tags on the dataset denoting the workbench server so I know which workbench server specifically this data came from. It's an extra step and I haven't found an easy way to automated it yet but it's been tremendously helpful to make sure we can track down the source of a dataset.

Another alternative is you can utilize a prefix on your datasets like [DEV] or [PROD] so you know one is a development dataset and the other is a production dataset.