- Remove CLEANUP_TODO.md and MIGRATION_SCAFFOLDING.md (internal notes)
- Remove temporary scripts: copy_and_run.sh, work.sh
- Remove AI assistant utility scripts: backup-claude-md.sh, codex-router.sh
These files were used during development but don't belong in the repository.
Implements #750 - allows hiding the username/password login form when
using OIDC SSO to avoid user confusion, while maintaining security.
- Added HideLocalLogin config option (env: PULSE_AUTH_HIDE_LOCAL_LOGIN)
- Exposed hideLocalLogin in /api/security/status endpoint
- Updated Login.tsx to conditionally hide local login form
- Added escape hatch via ?show_local=true URL parameter
This approach avoids the security and upgrade issues that led to
DISABLE_AUTH being removed (see #707, #678), while solving the UX
problem of users being confused by multiple login options.
- Allow HEAD requests in addition to GET for all download handlers
(install scripts, binaries, checksums) to prevent 405 errors
- Add /uninstall-host-agent.sh to special routes in ServeHTTP
- Add test coverage for HEAD request handling
- Resolves 'method not allowed' errors during agent installation
- Change default server listen addresses to empty string (listen on all interfaces including IPv6)
- Add short hostname matching fallback in host lookup API to handle FQDN vs short name mismatches
- Implement retry loop (30s) in both Windows and Linux/macOS installers for registration verification
- Fix lint errors: remove unnecessary fmt.Sprintf and nil checks before len()
This resolves the 'Installer could not yet confirm host registration with Pulse' warning
by addressing timing issues, hostname matching, and network connectivity.
- Add /download/pulse-host-agent.sha256 route to serve checksums
- Fixes 'Checksum not available' warning during host agent installation
- Bump version to 4.32.6
Related to #746
Introduced sync.RWMutex to protect concurrent access to configuration
fields (AuthUser, AuthPass, APITokens) that are modified by the
ConfigWatcher at runtime.
- Added global config.Mu RWMutex in internal/config/config.go
- Protected config updates in ConfigWatcher.reloadConfig() and reloadAPITokens()
- Protected config reads in CheckAuth and all API token handlers
- Protected Router.SetConfig() during full config reloads
This prevents race conditions when .env file changes trigger config
reloads while authentication handlers are reading the same fields.
- Removed global legacySSHDisabled flag that was triggered by any single node auth failure
- Changed disableLegacySSHOnAuthFailure to only log warnings
- Fixed potential context leak in monitor.go
- Updated tests to reflect removal of global disable logic
- Add fallback to project root scripts/ directory for install-docker-agent.sh
- Add fallback to project root bin/ directory for pulse-docker-agent binary
- Fixes 404 errors when downloading agent installer and binary in dev mode
- Production paths remain unchanged (/opt/pulse/...)
GitHub's API has a quirk: if you POST to /releases with a tag_name
that already exists as a git tag, it creates an 'untagged' release
instead of attaching to the existing tag.
The fix is to let the API create both tag and release together.
gh release create doesn't work properly when the tag already exists -
it creates an 'untagged' release instead of attaching to the existing tag.
Using the API directly with POST to /releases fixes this.
gh release create with --target was still creating untagged releases.
The fix is to create and push the git tag explicitly first, then
create the release which will properly attach to the existing tag.
The workflow was broken because it expected a tag to exist but the
documented process never created one. This caused gh release create
to fail with 'untagged' releases.
Changes:
- Workflow now creates the tag using --target flag
- Simplified release creation logic (no retry loops needed)
- Removed confusing comment about 'tag already exists'
This fixes the fundamental issue where the workflow and documented
process were out of sync.
Fixes an issue where pvecm status output using decimal node IDs (e.g. '1' instead of '0x1') caused node discovery to fail. Added test case for this format.
Increases confidence score for PBS when receiving 401/403 responses to avoid unnecessary probing of other endpoints that trigger auth failure logs.
Fixes#741
- Refactor collector to support mocking
- Fix ZFS detection to support 'fuse.zfs' and case-insensitivity
- Add regression tests for ZFS dataset deduplication
Fixes#727. Previously, if temperature monitoring was enabled and a node wasn't found in ClusterEndpoints, the entire node processing was skipped. This change ensures we only skip temperature collection.
The grep pattern was looking for 'pulse-sensor-proxy' as a standalone
string, but the actual mount line contains paths like:
mp0: /run/pulse-sensor-proxy,mp=/mnt/pulse-proxy,replicate=0
This caused the removal logic to never execute, leaving the old mp
mount in place and preventing the migration to lxc.mount.entry format.
Changed pattern to match either path component:
- /pulse-sensor-proxy (source path)
- /mnt/pulse-proxy (mount point)
Also removed space after colon in pattern to match actual format.
This completes the fix for temperature proxy setup on LXC containers.
