Workstream's integrations with dbt Cloud and Monte Carlo allow you to automatically present status pages for an asset directly within its asset details, as well as when viewing it through the Concierge. You also have the option to build manual status pages if your team becomes aware of a data issue. This gives you immediate insight into data quality and freshness, allowing your users to more easily recognize issues in your environment.
To build automatic status pages, you will need to connect either dbt Cloud or Monte Carlo:
How do status pages display?
There are a few ways that status pages call your attention to broken assets. The most detailed is on the asset detail page, where you can select the "asset status" tab from the right rail.
The status page is broken down into two sections, one that shows the overall status and uptime and another that shows past incidents. In each section we provide statuses for each of the models or sources that are relevant to the asset.
You can expand the status to show which tests identified the failure, allowing you to hone in on where the issue might have occurred.
This same status page is available when viewing the asset natively in your analytics tool with the Concierge browser extension. Simply expand the Concierge and locate the asset status section from the right rail.
In order to alert your entire team to data issues, Workstream also displays an error message whenever a user accesses an asset that has a failed status.
This provides a quick notification without the need to click into the actual status page.
How does Workstream identify which dbt Cloud assets to show status pages for?
Please note: you must have dbt Cloud connected to Workstream at the account level before newly configured exposures will create status pages. Please follow the instructions here to connect.
One of the first things you will notice after integrating your dbt Cloud instance is that not all assets will have a status page. Only assets that are configured as exposures in dbt Cloud will display a status page. These are easily identifiable in Workstream because they contain a dbt logo alongside their native application's icon in the knowledge repository.
For exposures to be linked to an asset, they must be explicitly defined within the .yml file for your dbt project. You can read more about how to configure these in dbt's documentation. Whatever models and sources you define as the exposure's dependencies will be used by Workstream to construct the asset's status page. Correctly defining these dependencies is the most important step, as it allows Workstream to identify which assets have failing dependencies.
How does Workstream identify which Monte Carlo assets to show status pages for?
If you have not yet connected your Monte Carlo environment to Workstream, you can do so by following the steps in this article.
The process of identifying status pages for Monte Carlo is simpler than for dbt. Once the integration has been connected, Workstream will automatically build status pages for any assets where monitoring is occurring in Monte Carlo.
Note: currently only Looker assets are supported in our Monte Carlo integration. We will add more as we are able.
Any asset with a Monte Carlo status page will display a Monte Carlo logo alongside the asset type when viewing the asset card in Workstream.
Where can I quickly see all failing assets?
In addition to seeing status pages within an asset's details, Workstream also gives you the ability to check out all asset that have had failures. You can locate this list by navigating to the "Insights" option from the left nav.
This will take you to the environment insights page. The "Total failing assets" section indicates how many assets currently have issues with data quality or freshness. Click on this section to view a list of all failing assets.
The number of failing assets is also available as part of our trends summary email, which can be set to send daily or weekly. We hope that this improved insight will give you more clarity on the health of your data, and help you catch issues before they impact your stakeholders.
How can I create status pages and incidents manually?
If an asset is not configured as an exposure in dbt Cloud and does not have a status page created by Monte Carlo, it will not by default have a status page attached. However, workspace owners have the ability to manually create status pages and incidents if they are aware of data issues that would not be detected automatically through one of our integrations.
To create a status page, head to the "asset status" tab in the asset's detail. Then find the "Create status page" button.
Once an asset's status page has been created, you have the option to create a new incident using the "+" button.
When creating a new incident, you will need to write in an incident name, establish a timeframe for the incident (including whether it is ongoing), set a status, and provide a description. Select "Create incident" to add your incident to the asset status page.
Any new incidents or updates will then appear in the incident history.
Click "View incident report" to see more information on the incident, or if you wish to amend the status or description.
How long does it take for new exposures to sync in?
The time it takes for a new exposure to sync in and create a status page is dependent upon a few factors, including if you have recently connected your integrations, and the size of your dbt or Monte Carlo instance. In general, however, new exposures should display within a few minutes.
I have connected dbt Cloud and created exposures, is there anything else I need to do?
Nope! An active dbt Cloud connection and configured exposures should be all that is required for Workstream to build a status page.
How often are status pages updated?
We rely on webhooks, which for dbt are sent at the end of every job run. If your job run has completed, Workstream should update the status immediately. As a failsafe, we also sync in current statuses for all exposures via a daily catch-up sync.
Are there other types of data monitoring you'd like to see in Workstream? Let us know in the messenger widget on this page, or by emailing [email protected]. We'd love to hear from you.