Fixing up deletes a bit (#327)

Co-authored-by: FloatingGhost <hannah@coffee-and-dreams.uk>
Reviewed-on: AkkomaGang/akkoma#327

CHANGELOG.md

@@ -10,12 +10,18 @@ The format is based on [Keep a Changelog](https://keepachangelog.com/en/1.0.0/).
 - Config: HTTP timeout options, :pool\_timeout and :receive\_timeout
 - Added statistic gathering about instances which do/don't have signed fetches when they request from us
 - Ability to set a default post expiry time, after which the post will be deleted. If used in concert with ActivityExpiration MRF, the expiry which comes _sooner_ will be applied.
+- Regular task to prune local transient activities
+- Task to manually run the transient prune job (pleroma.database prune\_task)
 
 ## Changed
 - MastoAPI: Accept BooleanLike input on `/api/v1/accounts/:id/follow` (fixes follows with mastodon.py)
 - Relays from akkoma are now off by default
 - NormalizeMarkup MRF is now on by default
 - Follow/Block/Mute imports now spin off into *n* tasks to avoid the oban timeout
+- Transient activities recieved from remote servers are no longer persisted in the database
+
+## Upgrade Notes
+- If you have an old instance, you will probably want to run `mix pleroma.database prune_task` in the foreground to catch it up with the history of your instance.
 
 ## 2022.11
 

config/config.exs

@@ -569,7 +569,8 @@ config :pleroma, Oban,
     new_users_digest: 1,
     mute_expire: 5,
     search_indexing: 10,
-    nodeinfo_fetcher: 1
+    nodeinfo_fetcher: 1,
+    database_prune: 1
   ],
   plugins: [
     Oban.Plugins.Pruner,
@@ -577,7 +578,8 @@ config :pleroma, Oban,
   ],
   crontab: [
     {"0 0 * * 0", Pleroma.Workers.Cron.DigestEmailsWorker},
-    {"0 0 * * *", Pleroma.Workers.Cron.NewUsersDigestWorker}
+    {"0 0 * * *", Pleroma.Workers.Cron.NewUsersDigestWorker},
+    {"0 3 * * *", Pleroma.Workers.Cron.PruneDatabaseWorker}
   ]
 
 config :pleroma, :workers,
@@ -605,7 +607,8 @@ config :pleroma, :workers,
     new_users_digest: :timer.seconds(10),
     mute_expire: :timer.seconds(5),
     search_indexing: :timer.seconds(5),
-    nodeinfo_fetcher: :timer.seconds(10)
+    nodeinfo_fetcher: :timer.seconds(10),
+    database_prune: :timer.minutes(10)
   ]
 
 config :pleroma, Pleroma.Formatter,

docs/docs/administration/CLI_tasks/database.md

@@ -159,3 +159,23 @@ Change `default_text_search_config` for database and (if necessary) text_search_
     ```
 
 See [PostgreSQL documentation](https://www.postgresql.org/docs/current/textsearch-configuration.html) and `docs/configuration/howto_search_cjk.md` for more detail.
+
+## Pruning old activities
+
+Over time, transient `Delete` activities and `Tombstone` objects
+can accumulate in your database, inflating its size. This is not ideal. 
+There is a periodic task to prune these transient objects, 
+but on first run this may take a while on older instances to catch up
+to the current day.
+
+=== "OTP"
+
+    ```sh
+    ./bin/pleroma_ctl database prune_task
+    ```
+
+=== "From Source"
+
+    ```sh
+    mix pleroma.database prune_task
+    ```

lib/mix/tasks/pleroma/database.ex

@@ -110,6 +110,14 @@ defmodule Mix.Tasks.Pleroma.Database do
     end
   end
 
+  def run(["prune_task"]) do
+    start_pleroma()
+
+    nil
+    |> Pleroma.Workers.Cron.PruneDatabaseWorker.perform()
+    |> IO.inspect()
+  end
+
   def run(["fix_likes_collections"]) do
     start_pleroma()
 

lib/pleroma/activity/pruner.ex

@@ -0,0 +1,41 @@
+defmodule Pleroma.Activity.Pruner do
+  @moduledoc """
+  Prunes activities from the database.
+  """
+  @cutoff 30
+
+  alias Pleroma.Activity
+  alias Pleroma.Repo
+  import Ecto.Query
+
+  def prune_deletes do
+    before_time = cutoff()
+
+    from(a in Activity,
+      where: fragment("?->>'type' = ?", a.data, "Delete") and a.inserted_at < ^before_time
+    )
+    |> Repo.delete_all(timeout: :infinity)
+  end
+
+  def prune_undos do
+    before_time = cutoff()
+
+    from(a in Activity,
+      where: fragment("?->>'type' = ?", a.data, "Undo") and a.inserted_at < ^before_time
+    )
+    |> Repo.delete_all(timeout: :infinity)
+  end
+
+  def prune_removes do
+    before_time = cutoff()
+
+    from(a in Activity,
+      where: fragment("?->>'type' = ?", a.data, "Remove") and a.inserted_at < ^before_time
+    )
+    |> Repo.delete_all(timeout: :infinity)
+  end
+
+  defp cutoff do
+    DateTime.utc_now() |> Timex.shift(days: -@cutoff)
+  end
+end

lib/pleroma/object/pruner.ex

@@ -0,0 +1,31 @@
+defmodule Pleroma.Object.Pruner do
+  @moduledoc """
+  Prunes objects from the database.
+  """
+  @cutoff 30
+
+  alias Pleroma.Object
+  alias Pleroma.Delivery
+  alias Pleroma.Repo
+  import Ecto.Query
+
+  def prune_tombstoned_deliveries do
+    from(d in Delivery)
+    |> join(:inner, [d], o in Object, on: d.object_id == o.id)
+    |> where([d, o], fragment("?->>'type' = ?", o.data, "Tombstone"))
+    |> Repo.delete_all(timeout: :infinity)
+  end
+
+  def prune_tombstones do
+    before_time = cutoff()
+
+    from(o in Object,
+      where: fragment("?->>'type' = ?", o.data, "Tombstone") and o.inserted_at < ^before_time
+    )
+    |> Repo.delete_all(timeout: :infinity, on_delete: :delete_all)
+  end
+
+  defp cutoff do
+    DateTime.utc_now() |> Timex.shift(days: -@cutoff)
+  end
+end

lib/pleroma/web/activity_pub/activity_pub.ex

@@ -105,6 +105,23 @@ defmodule Pleroma.Web.ActivityPub.ActivityPub do
     end
   end
 
+  @unpersisted_activity_types ~w[Undo Delete Remove]
+  @impl true
+  def persist(%{"type" => type} = object, [local: false] = meta)
+      when type in @unpersisted_activity_types do
+    {:ok, object, meta}
+    {recipients, _, _} = get_recipients(object)
+
+    unpersisted = %Activity{
+      data: object,
+      local: false,
+      recipients: recipients,
+      actor: object["actor"]
+    }
+
+    {:ok, unpersisted, meta}
+  end
+
   @impl true
   def persist(object, meta) do
     with local <- Keyword.fetch!(meta, :local),

lib/pleroma/web/activity_pub/side_effects.ex

@@ -288,7 +288,6 @@ defmodule Pleroma.Web.ActivityPub.SideEffects do
 
   # Tasks this handles:
   # - Delete and unpins the create activity
-  # - Replace object with Tombstone
   # - Set up notification
   # - Reduce the user note count
   # - Reduce the reply count

lib/pleroma/web/mastodon_api/controllers/account_controller.ex

@@ -222,7 +222,6 @@ defmodule Pleroma.Web.MastodonAPI.AccountController do
       |> Maps.put_if_present(:language, Pleroma.Web.Gettext.normalize_locale(params[:language]))
       |> Maps.put_if_present(:status_ttl_days, params[:status_ttl_days], status_ttl_days_value)
 
-    IO.inspect(user_params)
     # What happens here:
     #
     # We want to update the user through the pipeline, but the ActivityPub

lib/pleroma/workers/cron/database_prune_worker.ex

@@ -0,0 +1,32 @@
+defmodule Pleroma.Workers.Cron.PruneDatabaseWorker do
+  @moduledoc """
+  The worker to prune old data from the database.
+  """
+  require Logger
+  use Oban.Worker, queue: "database_prune"
+
+  alias Pleroma.Activity.Pruner, as: ActivityPruner
+  alias Pleroma.Object.Pruner, as: ObjectPruner
+
+  @impl Oban.Worker
+  def perform(_job) do
+    Logger.info("Pruning old data from the database")
+
+    Logger.info("Pruning old deletes")
+    ActivityPruner.prune_deletes()
+
+    Logger.info("Pruning old undos")
+    ActivityPruner.prune_undos()
+
+    Logger.info("Pruning old removes")
+    ActivityPruner.prune_removes()
+
+    Logger.info("Pruning old tombstone delivery entries")
+    ObjectPruner.prune_tombstoned_deliveries()
+
+    Logger.info("Pruning old tombstones")
+    ObjectPruner.prune_tombstones()
+
+    :ok
+  end
+end

lib/pleroma/workers/search_indexing_worker.ex

@@ -14,11 +14,10 @@ defmodule Pleroma.Workers.SearchIndexingWorker do
   end
 
   def perform(%Job{args: %{"op" => "remove_from_index", "object" => object_id}}) do
-    object = Pleroma.Object.get_by_id(object_id)
-
     search_module = Pleroma.Config.get([Pleroma.Search, :module])
 
-    search_module.remove_from_index(object)
+    # Fake the object so we can remove it from the index without having to keep it in the DB
+    search_module.remove_from_index(%Pleroma.Object{id: object_id})
 
     :ok
   end

priv/repo/migrations/20221129105331_add_notification_activity_id_index.exs

@@ -0,0 +1,7 @@
+defmodule Pleroma.Repo.Migrations.AddNotificationActivityIdIndex do
+  use Ecto.Migration
+
+  def change do
+    create(index(:notifications, [:activity_id]))
+  end
+end

priv/repo/migrations/20221129110627_add_bookmarks_activity_id_index.exs

@@ -0,0 +1,7 @@
+defmodule Pleroma.Repo.Migrations.AddBookmarksActivityIdIndex do
+  use Ecto.Migration
+
+  def change do
+    create(index(:bookmarks, [:activity_id]))
+  end
+end

priv/repo/migrations/20221129110727_add_report_notes_activity_id_index.exs

@@ -0,0 +1,7 @@
+defmodule Pleroma.Repo.Migrations.AddReportNotesActivityIdIndex do
+  use Ecto.Migration
+
+  def change do
+    create(index(:report_notes, [:activity_id]))
+  end
+end

priv/repo/migrations/20221129112022_add_cascade_to_report_notes_on_activity_delete.exs

@@ -0,0 +1,19 @@
+defmodule Pleroma.Repo.Migrations.AddCascadeToReportNotesOnActivityDelete do
+  use Ecto.Migration
+
+  def up do
+    drop(constraint(:report_notes, "report_notes_activity_id_fkey"))
+
+    alter table(:report_notes) do
+      modify(:activity_id, references(:activities, type: :uuid, on_delete: :delete_all))
+    end
+  end
+
+  def down do
+    drop(constraint(:report_notes, "report_notes_activity_id_fkey"))
+
+    alter table(:report_notes) do
+      modify(:activity_id, references(:activities, type: :uuid))
+    end
+  end
+end

priv/static/logo.svg

@@ -0,0 +1,53 @@
+<?xml version="1.0" encoding="UTF-8"?>
+<svg xmlns="http://www.w3.org/2000/svg" viewBox="0 0 794.02 449.34">
+  <defs>
+    <style>
+      .cls-1 {
+        fill: #fff;
+      }
+
+      .cls-2 {
+        fill: #2d2053;
+      }
+
+      .cls-3 {
+        fill: #462d7a;
+      }
+
+      .cls-4 {
+        stroke: #2c1e50;
+      }
+
+      .cls-4, .cls-5 {
+        stroke-miterlimit: 10;
+      }
+
+      .cls-5 {
+        stroke: #fff;
+      }
+
+      .cls-6 {
+        fill: #181127;
+      }
+    </style>
+  </defs>
+  <g id="Layer_2" data-name="Layer 2">
+    <path class="cls-3" d="M157.78,328.03c14.93,10.84,39.31-.17,41.84-1.23,17.72-7.43,29.58-23.57,49.21-50.87,13.76-19.14,15.81-25.23,20.91-24.94,15.15,.87,11.81,53.95,44.44,73.73,9.91,6.01,26.49,9.9,36.77,3.3,38.25-24.54,5.94-204.91-77.79-226.32-5-1.28-17.72-3.92-33.51,0-22.2,5.51-36.13,19.6-42.39,26.14-42.45,44.34-78.04,172.18-39.49,200.18Z"/>
+  </g>
+  <g id="Layer_7" data-name="Layer 7">
+    <path class="cls-2" d="M204.07,121.19c-1.95,2.08-2.59,2.62-4.05,4.07-3.72,3.69-6.99,6.64-7.27,7.66-2.34,8.62,150,55.54,152.87,47.02,.21-.62-.7-2.8-2.53-7.15,0,0-1.6-3.8-3.52-7.29-25.29-45.91-48.81-56.9-48.81-56.9-42.56-19.27-85.38,11.19-86.69,12.6Z"/>
+  </g>
+  <g id="Layer_9" data-name="Layer 9">
+    <path class="cls-4" d="M351.37,193.16c-5.77-11.54-85.59,16.83-154.76,27.39-21.09,3.22-38.13,4.31-47.3,4.75-.74,2.91-1.76,7.02-2.87,11.97-1.93,8.6-2.89,12.89-2.6,13.78,3.3,9.95,59.73-.88,99.18-7.64,32.67-5.6,115.14-18.96,114.61-30.77-.03-.69-1.11-4.01-3.27-10.65-1.78-5.47-2.67-8.2-2.98-8.83Z"/>
+  </g>
+  <g id="Layer_6" data-name="Layer 6">
+    <path class="cls-1" d="M253.58,138.31c-27.39-.52-46.38,38.21-37.98,54.55,10.09,19.62,65.5,18.26,74.77-3.3,7.21-16.78-11.38-50.77-36.79-51.24Z"/>
+  </g>
+  <g id="Layer_4" data-name="Layer 4">
+    <path d="M151,82.48c-6.55,27.74,252.45,113.97,267.56,89.66,9.24-14.87-64.9-83.62-163.53-97.57-39.06-5.52-100.95-5.14-104.03,7.91Z"/>
+  </g>
+  <g id="Layer_5" data-name="Layer 5">
+    <path class="cls-5" d="M221.03,89.73c.41-5.25,6.51-5.74,28.85-19.42,26.97-16.51,28.85-22.38,56.86-40.83,30.07-19.81,48.46-31.94,54.82-26.61,9.72,8.15-25.18,43.33-21.31,99.35,.87,12.61,3.12,17.79-.86,23.01-18.25,23.95-120.07-13.68-118.35-35.5Z"/>
+    <path class="cls-6" d="M791.6,449.34c3.22,0,3.22-5,0-5s-3.22,5,0,5h0Z"/>
+  </g>
+</svg>

test/pleroma/activity/pruner_test.exs

@@ -0,0 +1,27 @@
+defmodule Pleroma.Activity.PrunerTest do
+  use Pleroma.DataCase, async: true
+
+  alias Pleroma.Activity
+  alias Pleroma.Activity.Pruner
+
+  import Pleroma.Factory
+
+  describe "prune_deletes" do
+    test "it prunes old delete objects" do
+      user = insert(:user)
+
+      new_delete = insert(:delete_activity, type: "Delete", user: user)
+
+      old_delete =
+        insert(:delete_activity,
+          type: "Delete",
+          user: user,
+          inserted_at: DateTime.utc_now() |> DateTime.add(-31 * 24, :hour)
+        )
+
+      Pruner.prune_deletes()
+      assert Activity.get_by_id(new_delete.id)
+      refute Activity.get_by_id(old_delete.id)
+    end
+  end
+end

test/pleroma/object/pruner_test.exs

@@ -0,0 +1,41 @@
+defmodule Pleroma.Object.PrunerTest do
+  use Pleroma.DataCase, async: true
+
+  alias Pleroma.Delivery
+  alias Pleroma.Object
+  alias Pleroma.Object.Pruner
+
+  import Pleroma.Factory
+
+  describe "prune_deletes" do
+    test "it prunes old delete objects" do
+      new_tombstone = insert(:tombstone)
+
+      old_tombstone =
+        insert(:tombstone,
+          inserted_at: DateTime.utc_now() |> DateTime.add(-31 * 24, :hour)
+        )
+
+      Pruner.prune_tombstones()
+      assert Object.get_by_id(new_tombstone.id)
+      refute Object.get_by_id(old_tombstone.id)
+    end
+  end
+
+  describe "prune_tombstoned_deliveries" do
+    test "it prunes old tombstone deliveries" do
+      user = insert(:user)
+
+      tombstone = insert(:tombstone)
+      tombstoned = insert(:delivery, object: tombstone, user: user)
+
+      note = insert(:note)
+      not_tombstoned = insert(:delivery, object: note, user: user)
+
+      Pruner.prune_tombstoned_deliveries()
+
+      refute Repo.get(Delivery, tombstoned.id)
+      assert Repo.get(Delivery, not_tombstoned.id)
+    end
+  end
+end

test/pleroma/web/activity_pub/activity_pub_test.exs

@@ -8,6 +8,7 @@ defmodule Pleroma.Web.ActivityPub.ActivityPubTest do
 
   alias Pleroma.Activity
   alias Pleroma.Builders.ActivityBuilder
+  alias Pleroma.Web.ActivityPub.Builder
   alias Pleroma.Config
   alias Pleroma.Notification
   alias Pleroma.Object
@@ -2613,4 +2614,28 @@ defmodule Pleroma.Web.ActivityPub.ActivityPubTest do
     {:ok, user} = ActivityPub.make_user_from_ap_id("https://princess.cat/users/mewmew")
     assert user.name == " "
   end
+
+  describe "persist/1" do
+    test "should not persist remote delete activities" do
+      poster = insert(:user, local: false)
+      {:ok, post} = CommonAPI.post(poster, %{status: "hhhhhh"})
+
+      {:ok, delete_data, meta} = Builder.delete(poster, post)
+      local_opts = Keyword.put(meta, :local, false)
+      {:ok, act, _meta} = ActivityPub.persist(delete_data, local_opts)
+      refute act.inserted_at
+    end
+
+    test "should not persist remote undo activities" do
+      poster = insert(:user, local: false)
+      liker = insert(:user, local: false)
+      {:ok, post} = CommonAPI.post(poster, %{status: "hhhhhh"})
+      {:ok, like} = CommonAPI.favorite(liker, post.id)
+
+      {:ok, undo_data, meta} = Builder.undo(liker, like)
+      local_opts = Keyword.put(meta, :local, false)
+      {:ok, act, _meta} = ActivityPub.persist(undo_data, local_opts)
+      refute act.inserted_at
+    end
+  end
 end

test/pleroma/web/mastodon_api/update_credentials_test.exs

@@ -226,7 +226,7 @@ defmodule Pleroma.Web.MastodonAPI.UpdateCredentialsTest do
     test "does not allow negative integers other than -1 for TTL", %{conn: conn} do
       conn = patch(conn, "/api/v1/accounts/update_credentials", %{"status_ttl_days" => "-2"})
 
-      assert user_data = json_response_and_validate_schema(conn, 403)
+      assert json_response_and_validate_schema(conn, 403)
     end
 
     test "updates the user's AKAs", %{conn: conn} do

test/support/factory.ex

@@ -233,7 +233,7 @@ defmodule Pleroma.Factory do
     %Pleroma.Object{data: Map.merge(data, %{"type" => "Article"})}
   end
 
-  def tombstone_factory do
+  def tombstone_factory(attrs) do
     data = %{
       "type" => "Tombstone",
       "id" => Pleroma.Web.ActivityPub.Utils.generate_object_id(),
@@ -244,6 +244,7 @@ defmodule Pleroma.Factory do
     %Pleroma.Object{
       data: data
     }
+    |> merge_attributes(attrs)
   end
 
   def question_factory(attrs \\ %{}) do
@@ -520,6 +521,33 @@ defmodule Pleroma.Factory do
     |> Map.merge(attrs)
   end
 
+  def delete_activity_factory(attrs \\ %{}) do
+    user = attrs[:user] || insert(:user)
+    note_activity = attrs[:note_activity] || insert(:note_activity, user: user)
+
+    data_attrs = attrs[:data_attrs] || %{}
+    attrs = Map.drop(attrs, [:user, :data_attrs])
+
+    data =
+      %{
+        "id" => Pleroma.Web.ActivityPub.Utils.generate_activity_id(),
+        "type" => "Delete",
+        "actor" => note_activity.data["actor"],
+        "to" => note_activity.data["to"],
+        "object" => note_activity.data["id"],
+        "published" => DateTime.utc_now() |> DateTime.to_iso8601(),
+        "context" => note_activity.data["context"]
+      }
+      |> Map.merge(data_attrs)
+
+    %Pleroma.Activity{
+      data: data,
+      actor: data["actor"],
+      recipients: data["to"]
+    }
+    |> Map.merge(attrs)
+  end
+
   def oauth_app_factory do
     %Pleroma.Web.OAuth.App{
       client_name: sequence(:client_name, &"Some client #{&1}"),
@@ -676,4 +704,14 @@ defmodule Pleroma.Factory do
     }
     |> Map.merge(params)
   end
+
+  def delivery_factory(params \\ %{}) do
+    object = Map.get(params, :object, build(:note))
+    user = Map.get(params, :user, build(:user))
+
+    %Pleroma.Delivery{
+      object: object,
+      user: user
+    }
+  end
 end