Error after updating to v0.21.0

[alspires]

Hi all,

I've been using Ragflow without any issues until I upgraded to v0.21.0, but after changing to this tag, I get the exception "psycopg2.ProgrammingError: invalid dsn: invalid connection option "max_retries"" every time our container starts. We are using an AWS hosted psql instance for DB transactions.

These are the logs:

2025-10-20 10:53:10,590 WARNING  49 Realtime synonym is disabled, since no redis connection.
2025-10-20 10:53:10,590 INFO     49 MAX_CONTENT_LENGTH: 134217728
2025-10-20 10:53:10,591 INFO     49 MAX_FILE_COUNT_PER_USER: 0
Traceback (most recent call last):
  File "/ragflow/.venv/lib/python3.10/site-packages/peewee.py", line 3232, in connect
    self._state.set_connection(self._connect())
  File "/ragflow/.venv/lib/python3.10/site-packages/playhouse/pool.py", line 76, in inner
    return fn(self, *args, **kwargs)
  File "/ragflow/.venv/lib/python3.10/site-packages/playhouse/pool.py", line 169, in _connect
    conn = super(PooledDatabase, self)._connect()
  File "/ragflow/.venv/lib/python3.10/site-packages/peewee.py", line 3973, in _connect
    conn = psycopg2.connect(**params)
  File "/ragflow/.venv/lib/python3.10/site-packages/psycopg2/__init__.py", line 121, in connect
    dsn = _ext.make_dsn(dsn, **kwargs)
  File "/ragflow/.venv/lib/python3.10/site-packages/psycopg2/extensions.py", line 167, in make_dsn
    parse_dsn(dsn)
psycopg2.ProgrammingError: invalid dsn: invalid connection option "max_retries"
During handling of the above exception, another exception occurred:
Traceback (most recent call last):
  File "/ragflow/api/ragflow_server.py", line 103, in <module>
    init_web_db()
  File "/ragflow/.venv/lib/python3.10/site-packages/peewee.py", line 3127, in inner
    with ConnectionContext(self.db):
  File "/ragflow/.venv/lib/python3.10/site-packages/peewee.py", line 3122, in __enter__
    self.db.connect()
  File "/ragflow/.venv/lib/python3.10/site-packages/playhouse/pool.py", line 121, in connect
    return super(PooledDatabase, self).connect(reuse_if_open)
  File "/ragflow/.venv/lib/python3.10/site-packages/peewee.py", line 3231, in connect
    with __exception_wrapper__:
  File "/ragflow/.venv/lib/python3.10/site-packages/peewee.py", line 3059, in __exit__
    reraise(new_type, new_type(exc_value, *exc_args), traceback)
  File "/ragflow/.venv/lib/python3.10/site-packages/peewee.py", line 192, in reraise
    raise value.with_traceback(tb)
  File "/ragflow/.venv/lib/python3.10/site-packages/peewee.py", line 3232, in connect
    self._state.set_connection(self._connect())
  File "/ragflow/.venv/lib/python3.10/site-packages/playhouse/pool.py", line 76, in inner
    return fn(self, *args, **kwargs)
  File "/ragflow/.venv/lib/python3.10/site-packages/playhouse/pool.py", line 169, in _connect
    conn = super(PooledDatabase, self)._connect()
  File "/ragflow/.venv/lib/python3.10/site-packages/peewee.py", line 3973, in _connect
    conn = psycopg2.connect(**params)
  File "/ragflow/.venv/lib/python3.10/site-packages/psycopg2/__init__.py", line 121, in connect
    dsn = _ext.make_dsn(dsn, **kwargs)
  File "/ragflow/.venv/lib/python3.10/site-packages/psycopg2/extensions.py", line 167, in make_dsn
    parse_dsn(dsn)
peewee.ProgrammingError: invalid dsn: invalid connection option "max_retries"
2025-10-20 10:53:10,595 DEBUG    49 Using selector: EpollSelector

Could anyone give me a hand in figuring this one out, please?

Thanks!

[dosubot[bot]]

This error is caused by Ragflow v0.21.0 introducing new database connection options, specifically 'max_retries' and 'retry_delay', which are not valid parameters for psycopg2 (the PostgreSQL driver). These options were incorrectly passed to psycopg2.connect, resulting in the "invalid connection option 'max_retries'" exception. The fix, implemented in a later update, is to use a custom database class that removes these parameters before connecting to PostgreSQL, mirroring the MySQL implementation.

To resolve this on v0.21.0, you can either remove 'max_retries' and 'retry_delay' from your PostgreSQL connection configuration, or upgrade to a version where the fix is present and these parameters are handled correctly. More details and the fix are described in this pull request.

To reply, just mention @dosu.

---

^{How did I do? Good | Irrelevant | Incorrect | Verbose | Hallucination | Report 🐛 | Other}    

[alspires]

@dosu just an additional question, how do I use that commit specifically? Do I need to do git checkout <commit> by "commit" being the one from the pull request you sent me?

[dosubot[bot]]

Yes, you can use git checkout <commit_hash> to switch to the specific commit from the pull request that contains the fix (for example, git checkout 5b2e5dd3348c2e0a8a0bd8886ccb5ca718a236e4 for the PR #10524 fix) reference. This will put you in a detached HEAD state, which is fine if you just want to run that code. After checking out, make sure to reinstall dependencies and rebuild your container or environment to apply the changes—see the setup steps in the docs for running from source or rebuilding Docker images reference reference.

If you want to make changes or keep working from that commit, you can create a new branch with git checkout -b my-fix-branch.

To reply, just mention @dosu.

---

^{How did I do? Good | Irrelevant | Incorrect | Verbose | Hallucination | Report 🐛 | Other}