fix(stats): leave our own host out of the peers

Favicon lookups for local accounts can add our own host to instances.
Also skip userinfo and IPv6 literals in the backfill migration.

changelog.d/stats-peers-from-instances.change

@@ -1 +1 @@
-Read the peer list from the instances table and update instance stats every 5 minutes instead of every minute; the old peers query was one of the most expensive regular queries
+Read the peer list from the instances table (hosts of remote actors, without our own host) and update instance stats every 5 minutes; the old peers query scanned all remote users every minute

lib/pleroma/stats.ex

@@ -67,7 +67,10 @@ defmodule Pleroma.Stats do
   def calculate_stat_data do
     # Hosts are unique in instances, and remote users add theirs. Scanning
     # all remote users for their domains was one of the costliest queries.
-    peers = Repo.all(from(i in Instance, select: i.host))
+    # Favicon lookups for local accounts can add our own host, so skip it.
+    peers =
+      from(i in Instance, where: i.host != ^local_host(), select: i.host)
+      |> Repo.all()
 
     domain_count = Enum.count(peers)
 
@@ -93,6 +96,9 @@ defmodule Pleroma.Stats do
     }
   end
 
+  # Read from the config: this also runs before the endpoint has started.
+  defp local_host, do: Pleroma.Config.get([Pleroma.Web.Endpoint, :url, :host])
+
   @spec get_status_visibility_count(String.t() | nil) :: map()
   def get_status_visibility_count(instance \\ nil) do
     if is_nil(instance) do

priv/repo/migrations/20261028000000_backfill_instances_from_users.exs

@@ -13,7 +13,8 @@ defmodule Pleroma.Repo.Migrations.BackfillInstancesFromUsers do
       INSERT INTO instances (host, inserted_at, updated_at)
       SELECT DISTINCT host, now(), now()
       FROM (
-        SELECT substring(ap_id from '^[A-Za-z][A-Za-z0-9+.-]*://([^/:?#]+)') AS host
+        -- Skips userinfo; IPv6 literals give NULL and are left out.
+        SELECT substring(ap_id from '^[A-Za-z][A-Za-z0-9+.-]*://(?:[^/?#@]*@)?([^/:?#\\[]+)') AS host
         FROM users
         WHERE local = false AND ap_id IS NOT NULL
       ) AS hosts

test/pleroma/repo/migrations/backfill_instances_from_users_test.exs

@@ -19,12 +19,14 @@ defmodule Pleroma.Repo.Migrations.BackfillInstancesFromUsersTest do
     insert(:user, local: false, ap_id: "https://new.example/users/a")
     insert(:user, local: false, ap_id: "https://new.example/users/b")
     insert(:user, local: false, ap_id: "https://ported.example:8443/users/c")
+    insert(:user, local: false, ap_id: "https://user@userinfo.example/users/d")
+    insert(:user, local: false, ap_id: "https://[2001:db8::1]/users/e")
     insert(:user, local: true)
 
     migration.up()
 
     hosts = Instance |> Repo.all() |> Enum.map(& &1.host) |> Enum.sort()
-    assert hosts == ["known.example", "new.example", "ported.example"]
+    assert hosts == ["known.example", "new.example", "ported.example", "userinfo.example"]
 
     assert %{unreachable_since: ^unreachable_since} = Repo.get_by(Instance, host: "known.example")
   end

test/pleroma/web/mastodon_api/controllers/instance_controller_test.exs

@@ -86,6 +86,8 @@ defmodule Pleroma.Web.MastodonAPI.InstanceControllerTest do
     # Peers come from the instances table, which remote users add to.
     insert(:instance, host: "peer1.com")
     insert(:instance, host: "peer2.com")
+    # Favicon lookups for local accounts can add our own host.
+    insert(:instance, host: Pleroma.Web.Endpoint.host())
 
     Pleroma.Stats.force_update()