Skip to content

[skip ci] fix: batch fixes for issues #2090, 2091, 2093, 2095, 2098, 2169, 2170, 2171, 2172, 2173#393

Open
echobt wants to merge 1 commit intomasterfrom
fix/bounty-batch-2090-2173
Open

[skip ci] fix: batch fixes for issues #2090, 2091, 2093, 2095, 2098, 2169, 2170, 2171, 2172, 2173#393
echobt wants to merge 1 commit intomasterfrom
fix/bounty-batch-2090-2173

Conversation

@echobt
Copy link
Contributor

@echobt echobt commented Jan 27, 2026

Summary

Batch fix for multiple bounty issues.

Issues Fixed

Changes

  • cortex-cli/src/main.rs: Added validate_locale_env() function to warn about invalid LC_ALL/LANG values
  • cortex-cli/src/upgrade_cmd.rs: Added better messaging for beta/nightly channels, BrokenPipe handling
  • cortex-cli/src/scrape_cmd.rs: Added BrokenPipe handling for piped output
  • cortex-cli/src/export_cmd.rs: Added fsync() and file size verification for network filesystem reliability
  • cortex-app-server/src/config.rs: Added config file watcher that watches parent directory for atomic file replacement
  • cortex-linux-sandbox/src/run_main.rs: Added OOM kill detection using cgroup v2 memory events
  • cortex-engine/src/session.rs: Sessions list now skips corrupted files with warnings instead of failing

Note: CI skipped for cost control. Test manually before merge.

…, 2171, 2172, 2173 [skip ci]

Fixes:
- #2090: Added LC_ALL/LANG locale validation with warning for invalid values
- #2091: Improved upgrade --channel beta/nightly message when no versions available
- #2093: Added BrokenPipe handling for upgrade --check when output is piped
- #2095: Added BrokenPipe handling for --debug scrape output when piped
- #2098: Already fixed (--dry-run flag already exists in run command)
- #2169: Added fsync and file size verification for export to detect network filesystem issues
- #2170: Already fixed (warn_if_ambiguous_model already warns about ambiguous model names)
- #2171: Added config file watcher support for atomic file replacement (directory-based watching)
- #2172: Added OOM kill detection for sandbox process exits (cgroup v2 support)
- #2173: Sessions list now gracefully handles corrupted session files instead of failing entirely
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant