From f009b75841f93682694099c029179cbbba561c75 Mon Sep 17 00:00:00 2001
From: davelopez <46503462+davelopez@users.noreply.github.com>
Date: Wed, 10 Apr 2024 15:06:11 +0200
Subject: [PATCH] Make sure `data_dir` is used for default result backend DB

---
 doc/source/admin/galaxy_options.rst         |  7 ++++---
 lib/galaxy/config/__init__.py               | 14 ++++++++++++++
 lib/galaxy/config/sample/galaxy.yml.sample  |  7 ++++---
 lib/galaxy/config/schemas/config_schema.yml |  4 ++--
 4 files changed, 24 insertions(+), 8 deletions(-)

diff --git a/doc/source/admin/galaxy_options.rst b/doc/source/admin/galaxy_options.rst
index 5cd5d4d2649f..a8a7066d2554 100644
--- a/doc/source/admin/galaxy_options.rst
+++ b/doc/source/admin/galaxy_options.rst
@@ -5154,9 +5154,10 @@
     only if you have setup a Celery worker for Galaxy and you have
     configured the `celery_conf` option below. Specifically, you need
     to set the `result_backend` option in the `celery_conf` option to
-    a valid Celery result backend URL. By default a SQLite database is
-    used for storing task results, please use a more robust backend
-    for production setups like Redis. For details, see
+    a valid Celery result backend URL. By default, Galaxy uses an
+    SQLite database at '<data_dir>/results.sqlite' for storing task
+    results. Please use a more robust backend for production setups
+    like Redis. For details, see
     https://docs.galaxyproject.org/en/master/admin/production.html#use-celery-for-asynchronous-tasks
 :Default: ``false``
 :Type: bool
diff --git a/lib/galaxy/config/__init__.py b/lib/galaxy/config/__init__.py
index ce8e5e8dc2b0..886da6328cea 100644
--- a/lib/galaxy/config/__init__.py
+++ b/lib/galaxy/config/__init__.py
@@ -1082,6 +1082,9 @@ def _process_config(self, kwargs: Dict[str, Any]) -> None:
             self.amqp_internal_connection = (
                 f"sqlalchemy+sqlite:///{self._in_data_dir('control.sqlite')}?isolation_level=IMMEDIATE"
             )
+
+        self._process_celery_config(kwargs)
+
         self.pretty_datetime_format = expand_pretty_datetime_format(self.pretty_datetime_format)
         try:
             with open(self.user_preferences_extra_conf_path) as stream:
@@ -1203,6 +1206,17 @@ def _load_theme(path: str, theme_dict: dict):
         else:
             _load_theme(self.themes_config_file, self.themes)
 
+    def _process_celery_config(self, kwargs):
+        self.celery_conf = kwargs.get("celery_conf")
+        if self.celery_conf:
+            result_backend = self.celery_conf.get("result_backend")
+            if result_backend:
+                # If the result_backend is the default SQLite database, we need to
+                # ensure that the correct data directory is used.
+                if "results.sqlite" in result_backend:
+                    result_backend = f"db+sqlite:///{self._in_data_dir('results.sqlite')}?isolation_level=IMMEDIATE"
+                self.celery_conf["result_backend"] = result_backend
+
     def _check_database_connection_strings(self):
         """
         Verify connection URI strings in galaxy's configuration are parseable with urllib.
diff --git a/lib/galaxy/config/sample/galaxy.yml.sample b/lib/galaxy/config/sample/galaxy.yml.sample
index b379fa67e838..a7ed80d5a489 100644
--- a/lib/galaxy/config/sample/galaxy.yml.sample
+++ b/lib/galaxy/config/sample/galaxy.yml.sample
@@ -2761,9 +2761,10 @@ galaxy:
   # only if you have setup a Celery worker for Galaxy and you have
   # configured the `celery_conf` option below. Specifically, you need to
   # set the `result_backend` option in the `celery_conf` option to a
-  # valid Celery result backend URL. By default a SQLite database is
-  # used for storing task results, please use a more robust backend for
-  # production setups like Redis. For details, see
+  # valid Celery result backend URL. By default, Galaxy uses an SQLite
+  # database at '<data_dir>/results.sqlite' for storing task results.
+  # Please use a more robust backend for production setups like Redis.
+  # For details, see
   # https://docs.galaxyproject.org/en/master/admin/production.html#use-celery-for-asynchronous-tasks
   #enable_celery_tasks: false
 
diff --git a/lib/galaxy/config/schemas/config_schema.yml b/lib/galaxy/config/schemas/config_schema.yml
index 1f742f869339..8e9c0e44a13e 100644
--- a/lib/galaxy/config/schemas/config_schema.yml
+++ b/lib/galaxy/config/schemas/config_schema.yml
@@ -3764,8 +3764,8 @@ mapping:
           Activate this only if you have setup a Celery worker for Galaxy and you have
           configured the `celery_conf` option below. Specifically, you need to set the
           `result_backend` option in the `celery_conf` option to a valid Celery result
-          backend URL. By default a SQLite database is used for storing task results, please
-          use a more robust backend for production setups like Redis.
+          backend URL. By default, Galaxy uses an SQLite database at '<data_dir>/results.sqlite' for storing task results.
+          Please use a more robust backend for production setups like Redis.
           For details, see https://docs.galaxyproject.org/en/master/admin/production.html#use-celery-for-asynchronous-tasks
           
       celery_conf: