A regular try/catch block on its own won’t work as expected with Tasks. Internally runTask() throws some special errors to control flow execution. This is necessary to deal with resumability, serverless timeouts, and retrying Tasks.

If this function returns true you must rethrow the error. If it returns false you can handle the error as you normally would.

​
How to catch errors inside run()

You have two options:

  1. Use a regular try/catch block and isTriggerError() as documented on this page.
  2. Use the io.try() function.

​
Parameters

errorrequired

The error to check.

​
Returns

booleanrequired

true if the error is a Trigger Error, false otherwise.

You must rethrow the error if this function returns true.

client.defineJob({
  id: "get-repo-info",
  name: "GitHub get repo info",
  version: "0.1.0",
  integrations: { github },
  trigger: eventTrigger({
    name: "repo.info",
    schema: z.object({
      owner: z.string(),
      repo: z.string(),
    }),
  }),
  run: async (payload, io, ctx) => {
    //try
    try {
      return await io.github.getRepo("get-repo", {
        owner: payload.owner,
        repo: payload.repo,
      });
    } catch (error) {
      //if it is a TriggerError you must rethrow the error
      if (isTriggerError(error)) {
        throw error;
      }

      //do whatever you want when an error occurs
      return {
        success: false as const,
        error,
      };
    }
  },
});

Was this page helpful?

ConnectionAuthredactString