Deployment | dbt Labs

class: title, center, middle
# Deploying your dbt project

---

# We've built our first models... now what?

- How will our team members access them? They are currently in a schema named `dbt_jbloggs`

- What happens if our underlying data gets updated? (It should!)

- What happens if we add new models, or change existing ones?

---

# Solution:

- Run dbt on a schedule
- Use a "trusted" schema for this run

Also known as:
- Running dbt in production
- Deploying your dbt project

---

<h1>Deployment | Focus</h1>

<li> Understand why you need to deploy your dbt project </li>

<li> List some considerations when deploying your project </li>

<li> List some options for deploying your dbt project </li>

<li> Deploy your dbt project with dbt Cloud </li>

---
# Considerations

Does your production run:
* Use the latest version of your dbt project? → **continuous deployment**
* Use the correct dbt version for your project?
* Have monitoring and alerting if something goes wrong?
* Have a way to debug failures?
* Have a secure access model?
* Build & test your modified code every time you want to merge changes? →  **continuous integration**
* Have a secure way to host your project documentation?

---

# Deployment with dbt Cloud

The scheduler in dbt Cloud is purpose built for running dbt jobs on a schedule.

- ✅ Always pulls the latest version of your project
- ✅ Easy to build on Pull Request (and tear down the schema afterwards)
- ✅ Easy to generate and share documentation with users
- ✅ Secure access model
- ✅ Email and Slack notifications if something goes wrong, with rich logs
- ✅ Persist run artifacts for later comparison and analysis

---

# Scheduling a run in dbt Cloud

You'll need:
- A deployment environment
- A job, which includes:
    - Commands
    - A schedule

[Docs](https://docs.getdbt.com/docs/dbt-cloud/cloud-quickstart#create-a-new-job)

???
[teacher notes]
* there are two different types of environments: a deployment and a development.
* we've all been working in our own development environments, you each only have one.
* everyone in the project shares the same deployment environments. You can have as many as you need. 
* jobs are built in a specific deployment environment.

---

# The commands you should run as part of your dbt job

#### At a minimum:
- `dbt build`

#### Additionally:
- `Generate docs on run`
- `Run source freshness`

---

???
- This is what your Home Dashboard will look like once you have jobs running in Production.
- You'll see successes, errors, etc. and you can click through each invocation of a job for more information.

--
<img	
        style="width: 4%; position: absolute; bottom: 12%; right: 5%;"	
        src="/ui/img/dbt-standalone-new.png">
</img>	
<p style="width: 4%; position: absolute; bottom: 4%; right: 5.5%;">demo</p>

???
Demo Time:
- walk through development environment
- add deployment environment
- create a new job
- run job
- set up notifications for a job
- create a slim CI job
- mention environment variables and injecting logic into your SQL with jinja if blocks

---

# Additional strategies

Recommended:
- Separate "slim" CI job — fast runs, clear notifications
- Weekly full refresh job — more on this later!

Seen in the wild:
- Separate jobs for different parts of your project
- Blue/green deploys ([article](https://discourse.getdbt.com/t/performing-a-blue-green-deploy-of-your-dbt-project-on-snowflake/1349))
- Cloning from prod on CI runs (Snowflake only)

---

class: subtitle

# Knowledge check

You should be able to:
- Explain why you need to deploy your dbt project
- List some considerations when deploying your project
- List some options for deploying your dbt project
- Deploy your dbt project with dbt Cloud

---

# Resources

* [Deploy with dbt Cloud](https://docs.getdbt.com/docs/deploy/deployments)

* [Selection Syntax](https://docs.getdbt.com/reference/node-selection/syntax)

* [Model Timing](https://docs.getdbt.com/docs/deploy/dbt-cloud-job#model-timing)

* [Job Notifications](https://docs.getdbt.com/docs/deploy/job-notifications)

---
class: subtitle

#Zoom Out

<li>dbt Learn Norms</li>
<li>Why dbt?</li>
<li>Modeling</li>
<li>Sources</li>
<li>Docs</li>
<li>Deployment</li>