Help Centre › Building Processes
Real automations hit problems: a website is slow, a file is missing, an app shows an unexpected popup. Automize gives you two ways to handle these gracefully so a single failure doesn't halt the whole run – the exception connection and the Try-Catch block.
1. The exception connection
Every activity can have an EXCEPTION connection in addition to its normal one. If the step fails – an error, a timeout, something unexpected – the process follows the exception connection instead of the default one.
- Add the step you want to protect, and the step that should run if it goes wrong (for example "Log the error and send an alert").
- Draw a connection from the protected step to the recovery step.
- Click the connection and set its type to EXCEPTION. It is drawn in red on the map.
For more on connection types, see Working with Process maps.
2. The Try-Catch block
An exception connection protects a single step. A Try-Catch protects a whole group of steps at once. It is a container you place on the map with up to three zones:
- Try – the steps you want to run normally. The process attempts these in order.
- Catch – the steps that run if anything inside the Try zone fails. This is where you recover, log the problem, or notify someone.
- Finally (optional) – steps that run afterwards either way, whether the Try zone succeeded or failed. Use it for clean-up that must always happen, such as closing a file or logging out.
2.1 Add a Try-Catch block
- Open the Add panel and choose Try-Catch.
- Place the activities you want to protect inside the Try zone.
- Place your recovery activities inside the Catch zone.
- If you need guaranteed clean-up, turn on the Finally zone and add steps to it.
3. Which to use
- Use an exception connection for a single risky step with a simple recovery.
- Use a Try-Catch when several steps form one unit of work that should succeed or fail together, or when you need guaranteed clean-up with Finally.
When you test a process, failures and the paths taken are visible in the run log and history – see Running and debugging a process.