← All platforms

Elixir error tracking

The supervisor can restart the process. You’ll still want to know why it stopped. Use the Sentry Elixir SDK to keep those reports on your Telebugs server.

~/hoofnotes  nvim mix.exs config/config.exs check_notes.exs
defp deps do
[
{:sentry, "~> 13.0"},
{:jason, "~> 1.4"},
{:finch, "~> 0.21"}
]
end
import Config
config :sentry,
# Copy the complete Telebugs DSN, key and path included.
dsn: "YOUR_TELEBUGS_DSN",
environment_name: config_env()
try do
raise "Hello from Hoof Notes"
rescue
error ->
Sentry.capture_exception(error,
stacktrace: __STACKTRACE__, result: :sync)
end
~/hoofnotes  mix deps.get
~/hoofnotes  mix run check_notes.exs

Hang on, I already use Sentry

Keep sentry. Point it at your Telebugs DSN and keep reporting errors as before. The reports go to your server; you don’t need a Sentry account.

What about a process that crashes on its own?

The SDK’s logger integration can report OTP process crashes. For exceptions you rescue yourself, send the exception and its backtrace explicitly. Telebugs accepts both; the SDK integration determines what gets captured.

Does Phoenix fit here, too?

Yes. The Phoenix integration adds web request context and support for LiveView errors. It uses the same Sentry Elixir SDK.