Exploring the Blazer gem

Learn about the Blazer gem, the easiest way to query data from your Ruby on Rails apps.

It takes 2 minutes to install and lets you query your data in SQL, assemble queries into dashboards, run regular checks and much more.

I've been using — and loving — the Blazer gem a lot lately.

(To the point where I now auto-add it to any new side-project).

Blazer is perfect for querying your production database, pulling some stats, and whipping up a dashboard (without dancing the ssh-into-the-prod-server dance). It's basically a low-budget Metabase built specifically for Ruby on Rails.

It supports an astonishing number of data sources — 23 by my count — including classics like PostgreSQL, MySQL, SQLite, and BigQuery. It runs your queries in transactions (so you don't accidentally mutate data), and supports read-only users for extra safety. Oh, and Hacker News loves Blazer.

I've explored Blazer more below, broken down into:

• Writing Blazer queries,
• Creating Blazer dashboard
• Building Blazer checks
• And finally, getting started with Blazer

A typical Blazer query result

You'll probably spend about 90% of your time in Blazer crafting queries.

Here's Blazer's query page:

Your actual SQL query lives in the top-left, with the results below. Here, you can see Blazer's first trick: charts.

Blazer automatically generates charts based on the column types returned from your query.

In the example above, because we returned a string column followed by a numeric (count) column, Blazer rendered a handy bar chart.

Charts are one of my favorite Blazer features: they're easy to use (once you get the hang of them), and despite their, overall, low-budget feel, they're very handy.

Blazer has some official examples to try. Here's a basic query you can edit to test out it's interface.

Some more thoughts after writing plenty of Blazer queries:

• Forking & cloning queries is really neat & easy (blue fork button in the top-right group). I'll often use one query as the basis for another, tweak it until I'm happy, then fork it as it's own query. Blazer accommodates that workflow perfectly.
• The auto-generated charts can be fiddly to setup, since they depend on the order & typing of query results. I often flick back to the charts section of Blazer's README for help.
• I'd really, really love an "auto format SQL" button similar to the one in Metabase. It's such a handy QOL feature and without it, I find my queries get messy, fast (yes, I could just write neater queries 😅).

Wildcards in your queries

Blazer also supports {wildcard values} in your queries. Wildcards let you enter values on the fly rather than baking them into a query.

One of the official examples uses wildcards for start & end times. Blazer is even smart enough to render a calendar datepicker for easy input.

SELECT * FROM ratings
WHERE rated_at >= {start_time}
AND rated_at <= {end_time}

Another classic example is filtering records for a specific user_id, like:

select * from your_table
where user_id = {user_id}

Building dashboards from Blazer queries

Blazer lets you group multiple queries together into a dashboard (official example here). Handy, and self-explanatory.

Here's a screenshot from that dashboard example:

Dashboards work well — It's vastly easier to open 1 dashboard vs. N different queries.

I have a gripe though:

I find Blazer's dashboards quite visually sparse. Since each chart is stacked on top of each other, only ~1 result is visible at a time (the screenshot above is zoomed out).

It would be awesome to, one day, see a 2-column layout. You'd be able to fit more data into the same screen real estate. Of course, since Blazer is open source I could add that myself... we'll see 😅

Checks

Blazer also supports (wow it supports a lot of stuff, doesn't it?) Checks.

Checks are queries that run at set intervals, and automatically alert you if they fail.

Checks can fail based on these 3 conditions:

• Query returns no results,
• Query returns any results,
• Blazer detects an anomaly, backed by Blazer's anomaly detection integrations.

I won't comment here too much, since I haven't used Checks yet. But they seem handy, especially when coupled with the more sophisticated approaches to anomaly detection, like Prophet (another ankane gem!).

Blazer delivers checks via email or Slack. Again, very handy.

Getting started

So you've read this far: ready to give Blazer a try?

protip: Play around with Blazer locally in a git branch first

To get started with Blazer, follow these steps (from the official instructions):

# 1. install / add to Gemfile.rb
gem "blazer"

# 2. Generate the migrations
rails generate blazer:install
rails db:migrate

# 3. Mount Blazer to a path (in routes.rb)
mount Blazer::Engine, at: "blazer"

That installs everything and mounts Blazer to /blazer.

Next, add an environment variable to your production environment that points Blazer to your database:

ENV["BLAZER_DATABASE_URL"] = "postgres://user:password@hostname:5432/database"

Authentication is also a good idea, otherwise anyone could come along and query your database!

Basic authentication works fine, or you can hide Blazer behind Devise authentication:

# 1. Basic auth:
# add these ENV variables, or set them in an initializer.
#
ENV["BLAZER_USERNAME"] = "andrew"
ENV["BLAZER_PASSWORD"] = "secret"

# 2. Devise auth:
# You can also hide Blazer behind Devise.
#
authenticate :user, ->(user) { user.email == "you@example.com" } do
 mount Blazer::Engine, at: "blazer"
end

If you choose the basic auth option, you'll login to your Blazer dashboard like this:

It would be wise to also set up a read-only user for Blazer, although I tend to skip this step myself. I'd do this if multiple people were going to use Blazer though (it's easy to setup anyway: just a single command inside your database console).

Blazer also supports a basic permissions model which, again, would be worth exploring if you've got multiple people sharing Blazer.

That should be enough to get you started with Blazer though. Once you've explored it a bit (either locally or the online demo), make sure you read the full README, since it covers concepts I left out from this article.

Conclusion

If you couldn't already tell, I think Blazer is awesome.

It's a great way to get up-and-running with basic queries & dashboards in production, and saves you messing around with ssh and production consoles.

(It also pairs well with other gems like Ahoy, which I've written about further: Product analytics with the Ahoy gem (and Blazer))

At this point, it's probably saved me hours of re-writing the same queries and pulling the same data for basic things: new user signups, utilization of certain features etc.

Since Blazer's so well-integrated with Rails — and easy to install — it doesn't feel like overkill to mount it to a fresh Rails app with only a handful of users.