Repository navigation
Upgrading to the latest version corrupted logs_2.sqlite and codex app wouldn't restart #23917
Description
What version of the Codex App are you using (From “About Codex” dialog)?
Version 26.519.22136 (3003)
What subscription do you have?
Pro plan
What platform is your computer?
Darwin 25.4.0 arm64 arm
What issue are you seeing?
I just updated the app to the latest version, and it wouldn't relaunch, just gave me this error:
The app cannot finish launching until its SQLite database is accessible. Database path: Path unavailable in app-server startup error Close other Codex applications, then click Retry to check whether access has been restored. Error: (code=1, signal=null). Most recent error: Error: failed to initialize sqlite state runtime under /Users/latentflip/.codex: failed to initialize state runtime at /Users/latentflip/.codex: error returned from database: (code: 26) file is not a database
I tracked it down to logs_2.sqlite being corrupted somehow. I renamed that and things seem to have started okay now.
What steps can reproduce the bug?
N/A I can't.
What is the expected behavior?
Updating the app shouldn't corrupt sqlite files.
Additional information
No response
Activity
github-actions commentedon May 21, 2026
Potential duplicates detected. Please review them and close your issue if it is a duplicate.
Powered by Codex Action
etraut-openai commentedon May 21, 2026
The sqlite file was likely already corrupted prior to the upgrade. We were seeing intermittent reports of corruption with strange side effects. This version of the app now checks for corruption and reports it, which is why you just started to see this.
We're working on db auto-healing to eliminate this in most cases.
AS-1-Dev commentedon Aug 17, 2026
Hit this same failure mode on Windows and it reproduced the upgrade correlation described here — filed with full diagnostics as #39015.
~/.codex/logs_2.sqlitewas corrupt (quick_check:Rowid ... out of order,COUNT(*)→database disk image is malformed), which made the app unlaunchable withCodex app-server initialize handshake timed outon every attempt. Every other database inCODEX_HOMEpassedquick_checkcleanly, includingstate_5.sqlite.The upgrade correlation:
OpenAI.Codex26.707.8479.0 → 26.810.7004.0 (binaries dated 2026-08-16)logs_2.sqlite,-waland-shmaside together and the app rebuilds an empty log DBGiven this reproduced on a second machine and a different version pair, the upgrade path may be where the corruption is actually introduced rather than it being incidental. Would be worth reopening, or linking to #39015 if you'd rather track it there.