new ForEach topology show task in "parallel" while concurrency is set to 1

[Ben8t]

Describe the issue

When using the new ForEach task with concurency: 1 the topology still show tasks like if they were run in parallel. Not a big issue but could mislead users - especially those used to EachSequential + concurrency: 1 is the default value

Environment

• Kestra Version: 0.18-SNAPSHOT
• Operating System (OS/Docker/Kubernetes): Docker
• Java Version (if you don't run kestra in Docker):

[anna-geller]

I can confirm topology view is not correct

even though backend-wise everything is fine (see execution is correctly running each task groups sequentially one after the other rather than in parallel as shown in the topology view):

id: test_for_each
namespace: kestra.qa

tasks:
  - id: for_each
    type: io.kestra.plugin.core.flow.ForEach
    values: ['value1', 'value2', 'value3']
    concurrencyLimit: 1
    tasks:
      - id: log1
        type: io.kestra.plugin.core.log.Log
        message: "{{ taskrun.value }}"
      
      - id: log2
        type: io.kestra.plugin.core.log.Log
        message: "{{ taskrun.value }}"

      - id: log3
        type: io.kestra.plugin.core.log.Log
        message: "{{ taskrun.value }}"

@MilosPaunovic can you have a look? the tasks in the group should be shown as a sequential group within ForEach

[MilosPaunovic]

As I am far away from being an expert for this Topology/VueFlow implementation, there is a question for someone from the BE team to guide me a bit:

Right now the response of the endpoint returning the data for graph looks like this:

[
    // ...
    {
        "uid": "root.for_each.log1",
        "type": "io.kestra.core.models.hierarchies.GraphTask",
        "error": false,
        "task": {
            "id": "log1",
            "type": "io.kestra.plugin.core.log.Log",
            "message": "{{ taskrun.value }}"
        },
        "relationType": "PARALLEL"
    },
    {
        "uid": "root.for_each.log2",
        "type": "io.kestra.core.models.hierarchies.GraphTask",
        "error": false,
        "task": {
            "id": "log2",
            "type": "io.kestra.plugin.core.log.Log",
            "message": "{{ taskrun.value }}"
        },
        "relationType": "PARALLEL"
    },
    {
        "uid": "root.for_each.log3",
        "type": "io.kestra.core.models.hierarchies.GraphTask",
        "error": false,
        "task": {
            "id": "log3",
            "type": "io.kestra.plugin.core.log.Log",
            "message": "{{ taskrun.value }}"
        },
        "relationType": "PARALLEL"
    }
    // ...
]

Notice the "relationType": "PARALLEL" for the 3 tasks from flow Anna posted in a comment above. That is why the rendered output is parallel. Is this by design and should I try to manually handle it differently on the UI side or is there a need to change something on BE side also?

Possibly @loicmathieu could be the one who knows the answer to that?

[loicmathieu]

Yes this is a backend issue not a frontend issue.
The topology is driven by a graph computed by the backend