openai
Constant hang ups on "Working" or "Reconnecting"
What version of Codex is running?
0.47.0
What subscription do you have?
Pro
Which model were you using?
gpt-5-codex
What platform is your computer?
macOS
What issue are you seeing?
I operate with 3-5 codex sessions at a time in different projects. At least once a day, all my codex die at the same time. The only way to get them back to work, is to shut them all down, wait a minute, restart one, and say something simple like HELLO until it replies. When it does, I know I can continue. This has happened since I started using codex at least 2 months ago. It won't go away. It will say "reconnecting" forever with no fix in sight. I can't be the only one. It looks like something gets stuck, a session, a token for auth, something, that won't get the codexes in place to connect to the server. Until I shut it all down, and that is how it gets fixed.
What steps can reproduce the bug?
Uploaded thread: 019a12c4-551c-7683-9d75-d68377bc9014
- Use multiple codexes for 6+ hours
- All will stop working at one point
- Shut them all down, do nothing else
- Restart one
- Everything starts working again. No extra steps. No re-login. No cleanups. Nothing. Just just them down, and restart.
What is the expected behavior?
Codex continues to work for days without an issue!
Additional information
No response
Can you please run /feedback when one of instances them hangs up like that?
I just did... Its on 019a1815-637f-7d32-8943-0682da5b334e At least today it send me an error properly in the middle of the session ■ exceeded retry limit, last status: 401 Unauthorized, request id: 993c8e4c6a6b5cb2-RDU This is the FIRST time it has shown an actual error 401, never before, it would always hang.
Closing codex and re-opening solves the issue. No 401. Nothing.
Note: when reopening codex it jumped from 0.47.0 to 0.49.0
Well.. yeah, like I said, closing and opening codex works. You don't even have to do the /logout command. But that is not a fix, there's something really bad, because I have this happen twice a day, in all my codex instances. It's unusable. I come back to see a session has stopped working 1h ago or similar. It's ridiculous that a 401 problem and a refreshing JWT is causing this.
@pakrym-oai try this one: 9963bf64ab61c156-LIS, it's one of the failures. i had to logout in the VS Code extension and then login again to OpenAI
@Caibao7, what symptom are you seeing? Does the error message you're seeing refer to a 401 error, or is it a "Reconnecting" message?
Same problem.. reconnecting multiple times, then
unexpected status 400 Bad Request: { "error": { "message": "The encrypted content gAAA...8yU= could not be verified.", "type": "invalid_request_error", "param": null, "code": null } }
Same here.
{
"error": {
"message": "The encrypted content gAAA...Nli6 could not be verified.",
"type": "invalid_request_error",
"param": null,
"code": "invalid_encrypted_content"
}
}
@rmusser01, we're seeing that as well. The team is actively investigating it now.
Same issue there 019b4a6b-a231-7353-8a2b-3285530a8663. I'm experiencing it on every session heavily for a few days now.
This bug report has been open for a long time, and the posts cover many different issues, most of which have been resolved. I'm going to close and lock this bug report because it's not serving us well at this point.
@tryzang, you're likely hitting the issue that we're tracking in #8371. The team is actively working on that. If you run into this (or similar issue) again, please use the /feedback slash command to upload your session and post your thread ID.
@alanfloch, I see that you've also posted to #8371 already. Thanks!