fix(forgejo): actually install the pinned release (ETXTBSY on running binary) - #136
Merged
Merged
Conversation
… binary
setup.sh downloaded straight to /usr/local/bin/forgejo while forgejo.service
was running, which fails with ETXTBSY ("Text file busy"). Every deploy logged
"forgejo download failed" and prod stayed on 11.0.1 although the pin is
11.0.15. Download to forgejo.new, check its --version, and rename it over the
old binary; the existing `systemctl restart forgejo` then starts the new one.
A failed or wrong download now keeps the installed binary instead of warning
that the backend will be unavailable.
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
ThreatCrush Security Scan10 finding(s) HIGH/CRITICAL: 7 | LOW: 3
Snippets are redacted; ThreatCrush never prints matched credential material. |
ralyodio
added a commit
that referenced
this pull request
Oct 1, 2026
The redeploy restarts agentbbs, and the smoke test could run before it had re-bound :22 (0.3 s after on the #136 deploy). One probe made most deploys go red while the BBS answered moments later. Probe up to 12 times, 10 s apart, and fail only if :22 never answers. Still a keyscan only, so no account is ever registered. Co-authored-by: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Every deploy logs
forgejo download failed, so AgentGit is still on 11.0.1 (/api/v1/version→11.0.1+gitea-1.22.0) althoughFORGEJO_VERSIONpins 11.0.15.Cause:
setup.shcurls straight to/usr/local/bin/forgejowhileforgejo.serviceruns it, and Linux refuses to write a running executable (ETXTBSY, "Text file busy").Fix: download to
forgejo.new, check that its--versionmatches the pin, thenmvit over the old binary. A rename is allowed on a running executable. The existingsystemctl restart forgejo(same section) starts the new release. On a failed download the installed binary is kept, and the warning says so instead of claiming the backend is unavailable. This is the same patterndeploy/git-chovy/install.shuses.Effect on merge: the next deploy upgrades git.profullstack.com from 11.0.1 to 11.0.15 (same 11.0 LTS line, patch releases only) with one Forgejo restart.
Checks:
bash -npasses; shellcheck shows the same 3 pre-existing warnings as main, none new.Split out of #135 (closed): that PR's ufw/app.ini changes turned out to be unneeded, because :2222 is dropped by the DigitalOcean cloud firewall, not on the box.
🤖 Generated with Claude Code