Shared storages were appearing multiple times (once per node) because
the deduplication logic only checked the Proxmox `Shared` flag. Many
storage types are inherently cluster-wide but don't set this flag:
- RBD (Ceph block storage)
- CephFS
- PBS (Proxmox Backup Server)
- GlusterFS
- NFS
- CIFS/SMB
- iSCSI
Now we detect shared storage based on both the Shared flag AND the
storage type. Inherently shared storage types are deduplicated and
shown once with a "cluster" node designation.
Related to #1049
Add atomic `closed` flag to Client struct and `safeSend()` helper method
to prevent race condition when sending to client channels. The race
occurred when a client disconnected while a goroutine was trying to send
initial state - the channel could be closed between the registration
check and the actual send.
All sends to client.send now go through safeSend() which checks the
closed flag first. The flag is set atomically before closing the channel
in all code paths (unregister, dispatchToClients, broadcast, shutdown).
Related to #1048
When LoadActiveAlerts skipped acknowledged alerts older than 1 hour,
it was also not populating ackState. This meant that when the same
alert (e.g., backup-age) was recreated on the next poll cycle,
preserveAlertState couldn't find any acknowledgement record and
the alert would retrigger notifications.
Now ackState is populated even for skipped old acknowledged alerts,
so if they reappear, the acknowledgement will be restored.
Related to #1043
When --docker-runtime=podman is explicitly set, the agent should try
Podman-specific sockets first before falling back to environment
defaults (which try /var/run/docker.sock).
Also adds /var/run/podman/podman.sock as a candidate socket path,
which is used by CoreOS and some Fedora configurations.
Related to #1045
Changes:
1. Add MAX_POLL_TIMEOUT env var for large Proxmox clusters that need
more than 3 minutes for polling (default: 3m, minimum: 30s)
2. Handle external Ceph storage gracefully - don't mark nodes unhealthy
when Proxmox returns 'binary not installed' (e.g., for Ceph not
managed by Proxmox)
Related to #965
Previously, only PBS-type storages were queried when Active=0 because
querying inactive storage can return 500 errors. However, this caused
backups from datacenter backup tasks on shared storage (NFS, CIFS, etc.)
to not appear when the storage reported Active=0 on some nodes.
Now any storage with backup content is queried regardless of Active status.
If the storage is truly unavailable, GetStorageContent returns an error
which is already handled gracefully (logged and skipped).
Related to #1037
Update alerts for Docker containers are now available to all users,
not just Pro license holders. The feature alerts when container image
updates have been pending for longer than the configured delay
(default: 24 hours).
- Remove Pro license gating from update alerts
- Add FeatureUpdateAlerts to free tier features
- Remove obsolete license gating tests
Related to #1031
The test was checking an error path that no longer exists -
NewConfigPersistence now falls back to /etc/pulse when directory
creation fails, and calls log.Fatal() only when that also fails.
Tests were failing in CI because they created nodes.enc files without
valid encryption keys, triggering the crypto safety check.
Added createTestEncryptionKey helper and fixed:
- TestLoad_MockEnv (config_load_test.go)
- Multiple tests in commands_test.go that create nodes.enc
Test was creating .enc files without a valid encryption key, which
triggers the crypto safety check that prevents generating new keys
when encrypted data exists.
The blocked patterns for 'reboot' and 'shutdown' were too broad,
matching anywhere in the command string. This caused legitimate
Proxmox VM control commands like 'qm reboot 201' to be blocked
instead of requiring approval.
Fix by anchoring these patterns to only match bare system commands
(^reboot, ^shutdown, etc.) while allowing qm/pct variants through
the RequireApproval path.
Related to #1024
Tests were calling Load() without setting PULSE_DATA_DIR, causing them
to try to create /etc/pulse which fails in CI environments.
- Skip TestLoad_Defaults if /etc/pulse doesn't exist
- Add PULSE_DATA_DIR to tests that were missing it
Tests using /nonexistent/... paths fail in sandboxed environments
where they return 'permission denied' instead of 'not exists'.
Use /tmp/... paths instead which reliably return 'not exists'.
When a Proxmox cluster is discovered, Pulse now includes the user-provided
main host URL as a fallback endpoint. This handles scenarios where Proxmox
reports internal IPs that aren't reachable from Pulse's network (e.g.,
monitoring a remote cluster across different networks).
Previously, if all cluster endpoint IPs were unreachable, the connection
would fail with no fallback. Now the ClusterClient will fall back to the
main host URL, allowing Proxmox to route API calls internally.
Related to #1028
- Added TestPatrolService_RunPatrol_FullCoverage to test main patrol loop
- Added TestPatrolService_StartStop for lifecycle coverage
- Added TestPatrolService_Setters_Coverage for configuration setters
- Added TestPatrol_RunHeuristicAnalysis_Coverage for heuristic integration
- Mocked provider and state for deterministic AI patrol testing
- Addressed 0% coverage in internal/ai/patrol.go
- Fix TestMonitor_PollGuestSnapshots_Coverage by correctly initializing State ID fields
- Improve PBS client to handle alternative datastore metric fields (total-space, etc.)
- Add comprehensive test coverage for PBS polling, auth failures, and datastore metrics
- Add various coverage tests for monitoring, alerts, and metadata handling
- Refactor Monitor to support better testing of client creation and auth handling
When a Docker agent tries to register with a token that's already bound
to another agent, the error was logged generically as "Failed to send
docker report". Users had to dig into logs to understand the issue.
Now logs a prominent error message:
"DOCKER REGISTRATION FAILED: This API token is already used by another
Docker agent. Each Docker host requires its own unique token. Generate
a new token in Pulse Settings > Agents and reinstall with the new token."
Related to #1027
Reverse proxies (Traefik, Caddy, nginx) often normalize or reject URLs
containing %2F (encoded slash). Alert IDs contain forward slashes
(e.g., "docker-container-state-docker:abc/def"), causing acknowledge
requests to fail with 400 errors when going through a reverse proxy.
Added new body-based endpoints that accept alert ID in JSON body:
- POST /api/alerts/acknowledge {"id": "..."}
- POST /api/alerts/unacknowledge {"id": "..."}
- POST /api/alerts/clear {"id": "..."}
Updated frontend to use the new endpoints. Legacy path-based endpoints
are preserved for backwards compatibility.
Related to #1026
The mergeNVMeTempsIntoDisks and mergeHostAgentSMARTIntoDisks functions
require nodes to have LinkedHostAgentID populated to match disks with
host agent SMART data. However, the code was passing the local modelNodes
variable which doesn't have this field set - the linking happens inside
UpdateNodesForInstance which modifies the state's copy, not the local var.
Fixed by using currentState.Nodes (from GetSnapshot()) instead of
modelNodes/modelNodesCopy in both the skip-poll path and the background
goroutine. The state snapshot contains nodes with LinkedHostAgentID
already populated, allowing proper SMART data merging.
Related to #1014
When a container update command completed successfully, the server was
incorrectly returning shouldRemove=true, which caused the Docker host to
be removed and its API token revoked. This caused 401 Unauthorized errors
for subsequent agent reports.
The fix ensures shouldRemove is only true for "stop" commands, not for
"update_container" or "check_updates" commands.
Related to #1020
When collecting reports, the runtime re-detection was passing RuntimeAuto
instead of the user's configured preference. This caused podman to switch
back to docker on systems like CoreOS where podman provides a docker-
compatible socket at /var/run/docker.sock.
Now the current runtime (set at init from user's --docker-runtime flag)
is passed as the preference, preventing spurious runtime switching.
Related to #1022
The DisableDockerUpdateActions setting was being saved to disk but not
updated in h.config, causing the UI toggle to appear to revert on page
refresh since the API returned the stale runtime value.
Related to #1023
- Fix TestSaveHistoryWithRetry_WriteError to be robust on root
- Add TestOnAlert to history_test.go
- Add pmg_anomaly_test.go for PMG anomaly detection coverage
- Add cleanup_test.go for tracking map cleanup coverage
- extend filter_evaluation_test.go to cover all guest threshold logic
Ensures that LinkedHostAgentId, CommandsEnabled, IsLegacy, and LinkedNodeId
are correctly propagated to the frontend. This prevents regressions of the
bugs fixed for #952 and #971.
Related to #952 and #971
Both issues were caused by the backend not sending required fields to the
frontend in the ToFrontend() converters:
Issue #971 (Agent required badge):
- NodeFrontend was missing LinkedHostAgentId field
- Frontend couldn't identify linked host agents, so it fell back to showing
'Agent required' instead of 'Via agent'
Issue #952 (AI Commands toggle stuck):
- HostFrontend was missing CommandsEnabled field
- Frontend couldn't see the actual commandsEnabled state from the backend,
causing the optimistic UI to never receive confirmation that the state
had actually changed
Also added IsLegacy and LinkedNodeId to HostFrontend for completeness.
- Remove redundant nil checks before len() calls
- Mark unused parameters with underscore
- Convert if/else chains to switch statements for cleaner code
- Add test assertions to resolve unused write warnings in patrol_test.go
- Remove unused findAlertByID helper and its min dependency from update_alerts_test.go
- Remove redundant negative zero test case from utility_test.go (-0.0 == 0.0 in Go)
- Add lastSaveError, lastSaveTime, onSaveError fields to PatrolRunHistoryStore
- Add GetPersistenceStatus() and SetOnSaveError() methods
- Consistent with findings store and cost store error handling
- Add lastSaveError, lastSaveTime, onSaveError fields to cost.Store
- Add GetPersistenceStatus() method to check persistence health
- Add SetOnSaveError() callback for error notifications
- Rename scheduleSave to scheduleSaveLocked for clarity
- Document that scheduleSaveLocked must be called with lock held
- Add tests for new error tracking functionality
- Add Start/Stop lifecycle methods to AlertTriggeredAnalyzer
- Periodic cleanup of lastAnalyzed map every 30 minutes
- Prevents memory growth from stale cooldown entries
- Document that ai package feature constants are aliases of license constants
- Call Start() in StartPatrol and Stop() in StopPatrol
- Add tests for Start/Stop lifecycle
- Add missing KubernetesChecked field to persistence (data was being lost)
- Fix Duration field to properly convert between ms and nanoseconds
- Add automatic cleanup of stale stream subscribers (memory leak fix)
- Add error tracking for findings persistence with callback support
- Add GetPersistenceStatus() and SetOnSaveError() methods
- Add tests for new error tracking functionality
Implements PULSE_DISABLE_DOCKER_UPDATE_ACTIONS environment variable and
Settings UI toggle to hide Docker container update buttons while still
allowing update detection. This addresses requests for a 'read-only' mode
in production environments.
Backend:
- Add DisableDockerUpdateActions to SystemSettings and Config structs
- Add environment variable parsing with EnvOverrides tracking
- Expose setting in GET/POST /api/config/system endpoints
- Block update API with 403 when disabled (defense-in-depth)
Frontend:
- Add disableDockerUpdateActions to SystemConfig type
- Create systemSettings store for reactive access to server config
- Add Docker Settings card in Settings → Agents tab with toggle
- Show env lock badge when set via environment variable
UpdateButton improvements:
- Properly handle loading state (disabled + visual indicator)
- Use Solid.js Show components for proper reactivity
- Show read-only UpdateBadge when updates disabled
- Show interactive button when updates enabled
Closes discussion #982
This resolves issues where snapshots/backups persist after deletion if the
Instance field didn't match the ID prefix (due to case changes, name changes, etc).
Now consistent with how VMs, Containers, Storage, etc. are filtered.
Also adds Instance field to BackupTask model for completeness.
Addresses #1009 (refs #991)
This update integrates learned baselines into the heuristic analysis to detect abnormal behavior and records significant events (migrations, restarts, spikes) for correlation analysis. Also fixed syntax errors in Ollama integration tests.
This update integrates learned baselines into the heuristic analysis to detect abnormal behavior and records significant events (migrations, restarts, spikes) for correlation analysis.
- Add persistent volume mounts for Go/npm caches (faster rebuilds)
- Add shell config with helpful aliases and custom prompt
- Add comprehensive devcontainer documentation
- Add pre-commit hooks for Go formatting and linting
- Use go-version-file in CI workflows instead of hardcoded versions
- Simplify docker compose commands with --wait flag
- Add gitignore entries for devcontainer auth files
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
- Add DiskTemperature threshold to ThresholdConfig (default: 55°C trigger, 50°C clear)
- Process host SMART sensor data in CheckHost to generate disk_temperature alerts
- Add 'Disk Temp °C' column to Host Agents thresholds table in UI
- Make temperature tooltip interactive and scrollable to fix overflow issues
- Update AlertThresholds type to include diskTemperature field
Closes: #941
On errors, redirect back to login page with error params instead of
showing plain text error pages. This ensures users see friendly error
messages in the UI.
Related to #1006
rcourtman