Hydra ETL
Build your first job
Lesson 13 of 14 · Use containers

Organize workflow steps with containers

Use Sequence, Error, and Retry containers to group workflow steps, control failure behavior, and keep the canvas readable.

Containersabout 15 minutesworkflow configuration open

1. Objective

Goal

Find the Containers category, compare its three container types, place first-job inside each relevant scope, and learn how to collapse, detach, and configure the container from the workflow canvas.

Course project

Continue with studio-course, the manual first-workflow, and the saved first-job from the preceding lessons. Do not create a parallel replacement project.

Prerequisites

  • The top bar shows API online.
  • first-workflow is open in the editor.
  • first-job is available under Jobs in the node palette.

2. Steps

01Find the Containers category

Objective

Locate the grouping tools in the node palette.

Actions

  1. Open first-workflow.
  2. Stay on the current canvas while you scan the node palette.
  3. Find Containers (3) below Destinations.
NODE PALETTE → Containers
What you should see

The orange Containers (3) category appears below the workflow actions and control-flow nodes.

Hydra Studio canvas with the Containers category highlighted in the node palette
Find the category. It appears in the palette while you are still looking at the current job canvas.

02Compare the three container types

Objective

Choose a container from the behavior you need rather than from its appearance.

Actions

  1. Expand Containers (3).
  2. Read Sequence Container, Error Scope, and Retry Scope.
  3. Keep the category open for the next steps.
Containers (3) → expand
What you should see
Sequence Container  — group ordered steps
Error Scope         — choose behavior after child failure
Retry Scope         — retry a failed child
Expanded Containers category showing Sequence Container, Error Scope, and Retry Scope
Three containers, three purposes. Sequence organizes; Error and Retry add failure behavior.

03Group a job in a Sequence

Objective

Make first-job a child of a Sequence container and use its compact view.

Actions

  1. Select Workflow configuration.
  2. Drag Sequence Container onto empty canvas space.
  3. Drag first-job inside the orange frame.
  4. Release the job only when it is visibly contained.
  5. Select the chevron in the container header to collapse it, then expand it again.
Sequence Container ← first-job → collapse / expand
What you should see

The expanded container shows the job on the canvas. The collapsed container keeps a compact child list and displays a count of 1.

Expanded Sequence container with first-job placed inside it
Expanded Sequence. The job is visibly positioned inside the orange parent frame.
Collapsed Sequence container listing first-job and showing one child
Collapsed Sequence. The child remains attached and is listed in the compact container.

04Detach a step from its container

Objective

Remove the parent relationship without deleting either the job or the container.

Actions

  1. Expand the Sequence container.
  2. Open the context menu on first-job.
  3. Select Remove from container.
  4. Move the detached job to empty canvas space if needed.
first-job → More → Remove from container
What you should see

The job remains on the workflow canvas as an independent node. The Sequence container remains available and becomes empty.

Context menu for first-job with Remove from container highlighted
Detach, do not delete. Remove from container preserves the workflow step while removing its parent relationship.

05Choose an Error Scope policy

Objective

Control what the workflow does when a child inside the scope fails.

Actions

  1. Drag Error Scope onto the canvas.
  2. Move first-job inside the red frame.
  3. Open On child failure in the container header.
  4. Compare skip, continue, and fail, then select the behavior required by the workflow.
Error Scope ← first-job → On child failure
What you should see

The selected behavior appears in the container header and in the inspector's On failure field.

skip      — skip the failed child path
continue  — continue after the failure
fail      — fail the workflow
Error Scope containing first-job with the skip, continue, and fail policy menu open
Choose explicitly. The Error Scope exposes the failure policy directly above its child.

06Configure a Retry Scope

Objective

Retry a failed child a limited number of times with a delay between attempts.

Actions

  1. Drag Retry Scope onto empty canvas space.
  2. Move first-job inside the cyan frame.
  3. Set Retries to 3.
  4. Set delay to 5 seconds.
Retry Scope ← first-job → 3 retries × 5 s
What you should see

The Retry Scope header displays both values and the job remains visibly inside the scope.

Retries 3 ×, delay 5 s
Retry Scope containing first-job and configured for three retries with a five-second delay
Bound the retries. Both the attempt count and delay belong to the container that owns the child step.
Tips and traps

Use Sequence when the goal is organization, not failure recovery.

A node that merely overlaps a container may still be outside it. Confirm that it moves and collapses with the parent.

Use Remove from container to detach a child without deleting the workflow step.

Delete and Remove from container are different actions. Check the context-menu label before selecting it.

Choose an Error Scope policy deliberately: skip, continue, and fail produce different workflow outcomes.

Unlimited or aggressive retries can hide a persistent failure. Keep both the retry count and delay bounded.

3. Checklist

VerificationExpected result
PaletteSequence Container, Error Scope, and Retry Scope are visible
Sequencefirst-job stays attached in expanded and collapsed views
DetachmentThe job remains on the canvas after Remove from container
Error Scopeskip, continue, and fail are available
Retry ScopeThree retries and a five-second delay are visible

4. Troubleshooting

SymptomCauseFix
The job does not collapse with the SequenceIt overlaps the frame but is not a childDrag it out, then drop it again inside the expanded container
The job disappears after an actionDelete was selected instead of Remove from containerUndo the deletion, then use the job context menu to detach it
Error behavior is unexpectedThe wrong On child failure option is selectedOpen the Error Scope menu and confirm skip, continue, or fail
The retry takes too longThe retry count or delay is too highReduce the values in the Retry Scope header

5. Next lesson

The workflow now survives a failing step. One thing is still fixed in stone: every value it uses is written into a manifest. The last lesson moves those values onto the canvas, where a run can change them as it goes.

Close by

0 / 0 on this page