Fix recursion while waiting for redis connection

`Arq` has ability to retry connections to redis in case of problems.
That is a nice feature, but it crashes after hundreds of attempts
reaching maximal recursion depth. This change modifies re-try algorithm
from recursion to iteration avoiding the limit.

In practice it would be nice to have ability to wait forever as issue
with redis instance should not kill worker process, but that is a
separate change that can be built on top.

Fixes exception when retrying redis connection many times:
  RecursionError: maximum recursion depth exceeded while getting the str of an object

arq/connections.py

@@ -226,42 +226,34 @@ def pool_factory(*args: Any, **kwargs: Any) -> ArqRedis:
             ssl=settings.ssl,
         )
 
-    try:
-        pool = pool_factory(
-            db=settings.database, username=settings.username, password=settings.password, encoding='utf8'
-        )
-        pool.job_serializer = job_serializer
-        pool.job_deserializer = job_deserializer
-        pool.default_queue_name = default_queue_name
-        await pool.ping()
-
-    except (ConnectionError, OSError, RedisError, asyncio.TimeoutError) as e:
-        if retry < settings.conn_retries:
-            logger.warning(
-                'redis connection error %s:%s %s %s, %d retries remaining...',
-                settings.host,
-                settings.port,
-                e.__class__.__name__,
-                e,
-                settings.conn_retries - retry,
+    while True:
+        try:
+            pool = pool_factory(
+                db=settings.database, username=settings.username, password=settings.password, encoding='utf8'
             )
-            await asyncio.sleep(settings.conn_retry_delay)
+            pool.job_serializer = job_serializer
+            pool.job_deserializer = job_deserializer
+            pool.default_queue_name = default_queue_name
+            await pool.ping()
+
+        except (ConnectionError, OSError, RedisError, asyncio.TimeoutError) as e:
+            if retry < settings.conn_retries:
+                logger.warning(
+                    'redis connection error %s:%s %s %s, %d retries remaining...',
+                    settings.host,
+                    settings.port,
+                    e.__class__.__name__,
+                    e,
+                    settings.conn_retries - retry,
+                )
+                await asyncio.sleep(settings.conn_retry_delay)
+                retry = retry + 1
+            else:
+                raise
         else:
-            raise
-    else:
-        if retry > 0:
-            logger.info('redis connection successful')
-        return pool
-
-    # recursively attempt to create the pool outside the except block to avoid
-    # "During handling of the above exception..." madness
-    return await create_pool(
-        settings,
-        retry=retry + 1,
-        job_serializer=job_serializer,
-        job_deserializer=job_deserializer,
-        default_queue_name=default_queue_name,
-    )
+            if retry > 0:
+                logger.info('redis connection successful')
+            return pool
 
 
 async def log_redis_info(redis: Redis, log_func: Callable[[str], Any]) -> None:

tests/test_utils.py

@@ -29,6 +29,14 @@ async def test_redis_timeout(mocker, create_pool):
     assert arq.utils.asyncio.sleep.call_count == 5
 
 
+async def test_redis_timeout_and_retry_many_times(mocker, create_pool):
+    mocker.spy(arq.utils.asyncio, 'sleep')
+    retry_count = 2000
+    with pytest.raises(ConnectionError):
+        await create_pool(RedisSettings(port=0, conn_retry_delay=0, conn_retries=retry_count))
+    assert arq.utils.asyncio.sleep.call_count == retry_count
+
+
 @pytest.mark.skip(reason='this breaks many other tests as low level connections remain after failed connection')
 async def test_redis_sentinel_failure(create_pool, cancel_remaining_task, mocker):
     settings = RedisSettings()