Make cleaning up pushers depend on the device_id instead of the token_id (#15280)

This makes it so that we rely on the `device_id` to delete pushers on logout,
instead of relying on the `access_token_id`. This ensures we're not removing
pushers on token refresh, and prepares for a world without access token IDs
(also known as the OIDC).

This actually runs the `set_device_id_for_pushers` background update, which
was forgotten in #13831.

Note that for backwards compatibility it still deletes pushers based on the
`access_token` until the background update finishes.

1 files changed, 6 insertions, 1 deletions

diff --git a/synapse/push/__init__.py b/synapse/push/__init__.py
index a0c760239d..9e3a98741a 100644
--- a/synapse/push/__init__.py
+++ b/synapse/push/__init__.py
@@ -103,7 +103,7 @@ class PusherConfig:
 
     id: Optional[str]
     user_name: str
-    access_token: Optional[int]
+
     profile_tag: str
     kind: str
     app_id: str
@@ -119,6 +119,11 @@ class PusherConfig:
     enabled: bool
     device_id: Optional[str]
 
+    # XXX(quenting): The access_token is not persisted anymore for new pushers, but we
+    # keep it when reading from the database, so that we don't get stale pushers
+    # while the "set_device_id_for_pushers" background update is running.
+    access_token: Optional[int]
+
     def as_dict(self) -> Dict[str, Any]:
         """Information that can be retrieved about a pusher after creation."""
         return {