larshei

larshei

After doing the Ash Livebook Getting started for several times over the last 2? years,
I finally have a small project that I dare to tackle with Ash.

Loving it so far.

However, sometimes I am not quite sure what to do when receiving an error.

Basic Idea / Context

  1. The system is meant to collect “Proposals”.
  2. Proposals can be handed in, updated, accepted and rejected during a given timespan.
  3. Each given timespan is called a Proposal Round.
  4. All Proposals are related to a Proposal Round. They can only be interacted with while the Proposal Round is active.
  5. The Proposal Round being active is checked in a custom validation

The custom validation

defmodule Validations.OpenForProposals do
  use Ash.Resource.Validation

  @impl true
  # is this how we would properly check the round_id?
  def validate(%{attributes: %{round_id: id}}, _opts, _context) when not is_nil(id) do
    validate_open(id)
  end

  def validate(%{data: %{round_id: id}}, _opts, _context) when not is_nil(id) do
    validate_open(id)
  end

  def validate(_changeset, _opts, _context) do
    {:error, field: :round_id, message: "No proposal round was set."}
  end

  def validate_open(round_id) do
    with {:ok, round} <- Ash.get(Proposal, round_id),
         true <- round.is_open == true do
      IO.puts("OPEN FOR PROPOSALS") # ---- THIS IS PRINTED :) --------
      :ok
    else
      false -> {:error, field: :round_id, message: "Not currently open for proposals."}
      e -> e
    end
  end
end

Domain Module

defmodule Proposals do
  use Ash.Domain

  resources do
    resource Proposal do
      define :propose,
        args: [:employee_id, :round_id, :text],
        action: :propose

      define :accept,
        args: [:comment],
        action: :accept

      define :reject,
        args: [:comment],
        action: :reject
    end
  end
end

Resource Module

defmodule Proposal do
  use Ash.Resource, domain: Proposals, data_layer: AshPostgres.DataLayer

  actions do
    defaults [:read, :destroy, update: :*]

    create :propose do
      accept [:employee_id, :round_id, :text]

      validate {Validations.OpenForProposals, []}
    end

    update :accept do
      accept [:comment]

      validate {Validations.OpenForProposals, []}

      change set_attribute(:status, :accepted)
      change set_attribute(:comment, arg(:comment), set_when_nil?: false)
    end

    update :reject do
      accept [:comment]

      validate {Validations.OpenForProposals, []}

      change set_attribute(:status, :rejected)
      change set_attribute(:comment, arg(:comment), set_when_nil?: false)
    end
  end

  attributes do
    # ...
  end

  relationships do
    # ...
  end

  postgres do
    # ..
  end
end

The error

The current error is the following:

iex> {:ok, proposal} = Proposals.propose(me.id, 1, "Add 5 more chairs to the break room.")
  {:ok,
   #Proposal<
   ...
  >}
iex > Proposals.accept(proposal, "Good idea. Matthew will take care of it.")
  %Ash.Error.framework{
     errors: [
       %Ash.Error.Framework.MustBeAtomic{
         resource: Proposal,
         action: :accept,
         reason: "Validations.OpenForProposals does not implement `atomic/3`",
         splode: Ash.Error,
         bread_crumbs: [],
         vars: [],
         path: [],
         stacktrace: #Splode.Stacktrace<>,
         class: :framework
       }]}

So the validation works on :propose, but not on :accept.

I have read Basic bulk actions, atomics, new stream options, `error/2` expression and Update Actions — ash v3.29.3, but I do not understand what to do or what the error actually means (and why it occurs on the validation?!) :sweat_smile:

What am I not seeing / understanding and where I could I learn about it?

Showing Posts 1 to 7

03juan

03juan

As far as I can tell, in the info block for Fully Atomic Updates it states

An atomic update is one that can be done in a single operation in the data layer. This ensures that there are no issues with concurrent access to the record being updated, and that it is as performant as possible.

Since you’re fetching the record from the data layer and asserting on is_open in validate_open, there could be a race condition between application processes where one reads and acts on what it sees is a valid state in the data layer but before the write happens another process also changed the data so that the first process’s data is actually not valid any more.

So Ash is expecting you to tell it how the update could be done atomically using the atomic/3 callback.

See Validations without an atomic callback a little further down.

I think the first case works because it’s a create action, not an update

zachdaniel

zachdaniel

Creator of Ash

Exactly correct :slight_smile:

If this can’t be done atomically, you can add require_atomic? false to the action definition.

03juan

03juan

This is probably easier to do in the action with present/2

validate present([:round_id])

larshei

larshei OP

Alright, thanks!

in this scenario, race conditions do not really matter, cause all the steps are manual anyway, and the worst thing that could happen is someone submitting a proposal 5 ms after the round is being closed, which is not a big deal.

However, I have to admin I have no idea what I would need to do if race conditions actually mattered :smiley:

03juan

03juan

For that you’d have to figure out how to do a explicit blocking lock on the row, which would run your operation to validate the open status and then update the approval in a transaction. So any other processes are made to wait for the row to unlock before they can go on.

zachdaniel

zachdaniel

Creator of Ash

In Ash, this can be placed in your update action change get_and_lock(:for_update)

03juan

03juan

Beautiful. Love your work!

— All posts loaded —

Where Next? Top

Trending in Questions Top

water
I’m aware of 2 such projects, are there others? git: GitHub - team-alembic/realworld: A fullstack Phoenix LiveView application with back...
New
katta
I having some trouble figuring out if I have set myself too strict of standards for my production server. Currently I can handle 75% of r...
New
peck
Hi everyone. My team and I have been working on a fairly modest app based around video streaming and chat, but we’ve landed a customer t...
New
bradley
I really like the adapter patterns that ecto, nebulex, waffle, etc. use and would love find something similar for a key management servic...
New
unaware8150
Hello folks! So at work, we are seeing some situations where we have to define some “fixed” strings that are used across the codebase in...
New
Alvinkariuki
How Can I Optimise Compile Time Dependencies I have been building an elixir application for about 2 years now. Many modules and files ha...
New
leoschweizer
Hi, we run different LLM models across a cluster, with each node loading a configurable subset. Model-specific queues let nodes consume ...
New

Other Trending Topics Top

budgie
A little off-topic, but I feel like people here have a good head on their shoulders. I used to be quite good at making software. Was luc...
New
KristerV
Hey. Is there anyone here who creates agents in their apps? Not talking about using agents, but creating them. I’m finding it pretty diff...
New
mudasobwa
I fully migrated to my own harness from Anthropic/Gemini and I think it’s time to share it. Welcome DSH, the DeepSeek Harness, fully writ...
New
jimsynz
Beam Bots (or just BB for short) is a framework for building fault-tolerant robotics applications in Elixir using familiar OTP patterns. ...
New
juhalehtonen
There has been a thread to discuss the Stack Overflow Developer Survey on this forum every year since 2018, so here’s yet another one for...
New
budgie
I love Elixir. It’s one of 2 programming languages I’ve ever fallen in love with. But I don’t use it anymore. Serverless was the promis...
New

We're in Beta

About us Mission Statement

Options

Thread Display Mode




Thread Preview

Skip Thread Previews