›_ Codex Issues
Codex Issues / #15468

CLI: remote compact task fails with retry exhaustion and 429 Too Many Requests

Resolved 💬 2 comments Opened Mar 22, 2026 by araekan Closed Jun 29, 2026
💡 Likely answer: A maintainer (github-actions[bot], contributor) responded on this thread — see the highlighted reply below.

What version of Codex CLI is running?

0.116.0

What subscription do you have?

ChatGPT Plus

Which model were you using?

gpt-5.1-codex-mini

What platform is your computer?

Linux 6.1.0-44-arm64 aarch64

What terminal emulator and version are you using (if applicable)?

SSH from macOS Terminal.app

What issue are you seeing?

I am seeing a remote compact task failure in Codex CLI that ends with a rate-limit response after retries are exhausted.
■ Error running remote compact task: exceeded retry limit, last status: 429 Too Many Requests, request id: req_88e475f349d344c2a936778c88b0b02e
This appears to be happening during the remote compact/feedback flow rather than from a normal local shell command failure.

What steps can reproduce the bug?

  1. Start a Codex CLI session and work through a normal coding/debugging flow.
  2. Continue long enough for Codex to trigger a remote compact task.
  3. Observe the task fail after retries with:

• 429 Too Many Requests
• exceeded retry limit

  1. Codex then reports the uploaded feedback/thread ID and asks to open a GitHub issue.

Uploaded thread: 019ceeb4-4d71-77a1-9bb7-b02db3095a2c

What is the expected behavior?

The remote compact task should either:
• complete successfully, or
• fail gracefully with a clearer actionable message and recovery path.

It should ideally not hit repeated retries and then abort with a final 429 if the client can back off or defer the compact task more cleanly.

Additional information

_No response_

View original on GitHub ↗

2 Comments

github-actions[bot] contributor · 5 months ago

Potential duplicates detected. Please review them and close your issue if it is a duplicate.

  • ##15105

Powered by Codex Action

etraut-openai contributor · 1 month ago

Thanks for reporting this problem. Until recently, Codex used a separate endpoint and server-side logic for compaction. We recently switched to a more robust approach where the compaction logic is moved locally using the same endpoint as normal turns. This eliminates "remote compaction" errors and increases the stability and reliability of compaction operations. If you see further problems with compaction, please use /feedback and open a new bug report.