Content was being streamed twice:
1. During each iteration of the tool loop (intended for intermediate feedback)
2. Again after the loop ended with finalContent (redundant)
This caused duplicate responses when using Ollama and other providers.
- Add container update command handling to unified agent
- Agent can now receive update_container commands from Pulse server
- Pulls latest image, stops container, creates backup, starts new container
- Automatic rollback on failure
- Backup container cleaned up after 5 minutes
- Added comprehensive test coverage for container update logic
BREAKING CHANGE: AI Patrol now uses EXACT alert thresholds by default
instead of warning 5-15% before the threshold.
Changes:
- Default behavior: Patrol warns at your configured threshold (e.g., 96% = warns at 96%)
- New setting: 'use_proactive_thresholds' enables the old early-warning behavior
- API: Added use_proactive_thresholds to GET/PUT /api/settings/ai
- Backend: Added SetProactiveMode/GetProactiveMode to PatrolService
- Backend: Added GetThresholds to PatrolService for UI display
- Tests: Updated and added tests for both exact and proactive modes
- Also fixed unused imports in dockeragent/agent.go
When proactive mode is disabled (default):
- Watch: threshold - 5% (slight buffer)
- Warning: exact threshold
When proactive mode is enabled:
- Watch: threshold - 15%
- Warning: threshold - 5%
Related to #951
Docker deployments with custom port mappings would show incorrect URLs
in email alerts because the auto-detection couldn't determine the
external port.
Added a "Public URL" setting in Settings > Network that allows users
to configure the dashboard URL used in email notifications.
- Added publicURL field to SystemSettings (persistence.go)
- Load/save publicURL in system settings handler
- Apply publicURL to notification manager on change
- Added UI input in NetworkSettingsPanel
- Shows env override warning if PULSE_PUBLIC_URL is set
Related to #944
When a node was manually added with a hostname (e.g., pve.example.com)
and then the agent registered using its IP address, the code would
correctly deduplicate but incorrectly overwrite the user's configured
hostname with the agent's IP.
Now when matching by IP resolution (hostname resolves to agent's IP),
we preserve the user's original hostname configuration instead of
replacing it with the IP.
Related to #940
The GuestURL field was missing from NodeFrontend and its converter,
causing configured Guest URLs to be ignored when clicking on cluster
node names. The frontend would fall back to the auto-detected IP
instead of using the user-configured Guest URL.
Related to #940
When a powered-off VM is backed up by Proxmox, the VM status briefly
changes (e.g., to "running" during backup). This caused the powered-off
alert to be cleared, deleting the ackState record. When the backup
completed and the alert was recreated, it appeared as a new unacknowledged
alert, generating a new notification.
The fix preserves ackState when alerts are removed, allowing
preserveAlertState to restore the acknowledgement when the same alert
reappears. Stale ackState entries (for alerts that don't exist) are
cleaned up after 1 hour.
Related to #937
When GetWebhooks returns webhooks, headers and customFields are masked
with ***REDACTED*** for security. However, when the frontend toggled
a webhook's enabled state, it sent back the redacted values, which
overwrote the actual header values (like Authorization tokens).
This broke webhooks after disabling and re-enabling them, as the auth
headers were replaced with "***REDACTED***".
Now UpdateWebhook detects redacted values and preserves the original
headers/customFields from the existing webhook.
Related to #938
- Add registry checker tests (caching, enable/disable, parsing, concurrency)
- Add alert integration tests for update detection and Pro license gating
- Add API handler tests for /api/infra-updates endpoints
- Test cleanup of tracking maps when containers are removed
- Test threshold-based alerting behavior
- Add FeatureUpdateAlerts constant for Pro license gating
- Add feature to all Pro tier feature lists
- Add SetLicenseChecker method to alerts Manager
- Check Pro license in checkDockerContainerImageUpdate before alerting
- Wire license checker from router to alert manager
Free users still see update badges in the UI.
Pro users get proactive alerts after 24h of pending updates.
- Add GHCR (GitHub Container Registry) token support for public images
- Clean up dockerUpdateFirstSeen tracking when containers are removed
- Improve UpdateIcon tooltip to show digest info
- Add cursor-help to indicate hoverable tooltip
- Add routes for infrastructure update detection API:
- GET /api/infra-updates - list all container updates with filtering
- GET /api/infra-updates/summary - aggregated stats per host
- GET /api/infra-updates/host/{hostId} - updates for specific host
- GET /api/infra-updates/{resourceId} - specific resource update status
- POST /api/infra-updates/check - trigger update check (placeholder)
- Update handlers to query Docker container updates from monitor state
- Protected by auth and monitoring_read scope
When a Proxmox host has multiple network interfaces (management, Ceph,
cluster ring), the agent would use heuristic scoring to pick an IP,
which could select an isolated network instead of the management network.
Now the agent first determines which local IP is actually used to connect
to the Pulse server, ensuring registration uses a reachable IP. Falls back
to the heuristic scoring if connection-based detection fails.
Related to #929
When the same Ceph cluster is reported from multiple sources (PVE API
and host agent), it showed up twice in the UI. Now we deduplicate by
FSID before converting to frontend format, keeping the cluster entry
with the most complete data (most monitors/managers/pools reported).
Related to #928
Ubuntu was showing as "debian 24.04" because we used PlatformFamily
(which is "debian" for all Debian derivatives) instead of Platform
(which is "ubuntu" for Ubuntu).
Now uses Platform first, falling back to PlatformFamily only if empty.
Related to #927
When an agent registers using an IP address, check if any existing node's
hostname resolves to that same IP. This prevents duplicates when a node
was manually configured via hostname and later the agent is installed
which registers using the host's IP.
Changes:
- Add extractHostIP() to extract IP from URL if present
- Add resolveHostnameToIP() with 2s timeout for DNS resolution
- During agent auto-registration, check if existing hostname-based
configs resolve to the new IP and update instead of creating duplicates
- Add test for extractHostIP helper function
The legacy state file could represent either PVE or PBS registration,
depending on what was installed at the time. Now we check what's
currently installed to determine the correct behavior:
- If PVE is installed: legacy file means PVE was registered
- If PBS-only (no PVE): legacy file means PBS was registered
When PBS is installed directly on a PVE host (an officially supported
configuration), the agent now detects and registers BOTH products instead
of only detecting PVE.
Changes:
- Add detectProxmoxTypes() to detect all Proxmox products on a host
- Add RunAll() method to register each detected product separately
- Use per-type state files (proxmox-pve-registered, proxmox-pbs-registered)
to track registration status for each product independently
- Maintain backward compatibility with legacy single state file
- Add tests for new state file path logic
- Add DELETE /api/agents/unregister endpoint for agent self-unregistration
- Agent now unregisters itself from Pulse server when uninstalled
- Add clarifying note in UnifiedAgents explaining linked agents behavior
- Linked agents are managed via their PVE node but this is now explained in UI
- Add LastSeen field to HostAgent model for better agent status tracking
- Add /api/agents/host/uninstall endpoint for agent self-unregistration
- Update install.sh to notify server during --uninstall (reads agent ID from disk)
- Update install.ps1 with same logic for Windows
- Update frontend uninstall command to include URL/token flags
This ensures that when an agent is uninstalled, the host record is
immediately removed from Pulse and any linked PVE nodes have their
+Agent badge cleared.
When the unified agent binary isn't found locally (happens on LXC/barebone
installations that update via web UI which only updates the pulse binary),
redirect to GitHub releases using HTTP 307.
This complements the install.sh GitHub proxy fallback from 7b6613bb.
Related to #909
When install.sh or install.ps1 don't exist locally (happens on LXC/barebone
installations that were updated via web UI which only updates the binary),
fallback to fetching from GitHub raw content.
Related to #909
When nodes are updated, now validates that LinkedHostAgentID points to
an existing host agent. References to deleted host agents are automatically
cleared, fixing the 'Agent' tag persistence for users who removed agent
entries before commit c394d24.
Related to #920
When a host agent is deleted via the UI, the LinkedHostAgentID on any
PVE nodes that were linked to it was not being cleared. This caused
the "Agent" tag to persist in the UI after uninstalling the agent.
Related to #920
Acknowledged alerts were still triggering repeated webhook notifications
because the re-notification logic only checked cooldown period, not
acknowledgment status. Now acknowledged alerts are skipped entirely.
Related to #921
When adding a PBS node with username/password credentials, Pulse now
automatically:
1. Connects to PBS using the provided credentials
2. Creates a 'pulse-monitor@pbs' user with Audit permissions
3. Generates an API token
4. Stores the token instead of the password
This enables one-click PBS setup for Docker/containerized deployments
where you can't easily run the agent installer. Simply enter root@pam
credentials in the UI and Pulse handles the rest.
Falls back to password auth if token creation fails (e.g., old PBS
version or permission issues).
- Added PBS client methods: CreateUser, SetUserACL, CreateUserToken
- Added SetupMonitoringAccess() turnkey method that creates user + token
- Updated handleSecureAutoRegister to use PBS API for token creation
- Enables one-click PBS setup for Docker/containerized deployments
When users provide PBS root credentials, Pulse can now create the
monitoring user and API token remotely via the PBS API, eliminating
the need to SSH/exec into the container manually.
- Add Env field to Container struct in pkg/agents/docker/report.go
- Extract env vars from inspect.Config.Env in Docker agent
- Mask sensitive values (password, secret, key, token, etc.) with ***
- Display env vars in container drawer with green badges (amber for masked)
- Add tests for maskSensitiveEnvVars function
Related to #916
- patrol.go: Auto-fix now requires both config flag AND ai_autofix license
- service.go: IsAutonomous() checks for ai_autofix license before enabling
- ai_handlers.go: API returns 402 if enabling auto-fix/autonomous without license
- Add power.go with Intel RAPL and AMD energy driver support
- Read CPU package, core, and DRAM power consumption in watts
- Sample energy counters over 100ms interval to calculate power
- Add PowerWatts field to Sensors struct for API reporting
- Integrate power collection into host agent sensor gathering
- Add comprehensive tests for power collection module
Supports Intel CPUs (Sandy Bridge+) via RAPL and AMD Ryzen/EPYC
via the amd_energy kernel module.
Closescommunity-scripts/ProxmoxVE#9575
Extended lm-sensors parsing to capture all sensor readings:
- Fan speeds (RPM) from SuperIO chips like NCT6687
- Additional temperatures (DDR5/RAM, motherboard, etc.)
- All sensors not already captured as CPU/NVMe/GPU
Updated frontend tooltip to display fans and additional sensors
in separate sections with formatted names.
Closes discussion #911
Users can now exclude specific mount points from disk monitoring:
- Via CLI: --disk-exclude /mnt/backup --disk-exclude '/media/*'
- Via env: PULSE_DISK_EXCLUDE=/mnt/backup,*pbs*
Patterns support:
- Exact paths: /mnt/backup
- Prefix patterns: /mnt/ext*
- Contains patterns: *pbs*
This addresses the common case where external disks or
PBS datastores are being monitored but shouldn't be.
- Add smartctl package to collect disk temperature and health data
- Add SMART field to agent Sensors struct
- Host agent now runs smartctl to collect disk temps when available
- Backend processes agent SMART data for temperature display
- Graceful fallback when smartctl not installed
When two Proxmox nodes have the same hostname (e.g., 'px1' on different IPs),
the getHostAgentTemperature function was matching by hostname alone, causing
both nodes to show temperature from whichever host agent appeared first.
The fix:
- Added getHostAgentTemperatureByID that first tries matching by LinkedNodeID
(the unique node ID) before falling back to hostname matching
- Updated the caller to pass modelNode.ID for precise matching
- Maintains backwards compatibility for setups where linking hasn't occurred
Related to #891
When the server disables command execution for an agent, we now properly
call Close() on the command client to tear down the WebSocket connection.
Previously we just set the pointer to nil which left the goroutine running
with an orphaned connection.
This implements full remote configuration for the AI command execution setting:
Backend:
- Add CommandsEnabled field to HostMetadata for persistent storage
- Add GetHostAgentConfig/UpdateHostAgentConfig methods to Monitor
- Add /api/agents/host/{id}/config endpoint (GET for agents, PATCH for UI)
- Server includes config in report response for immediate agent application
- Agent parses response and dynamically enables/disables command client
Frontend:
- Add 'AI Commands' toggle column in Managed Agents table
- Toggle immediately updates server config; agent applies on next heartbeat
- Add 'Enable AI command execution' checkbox in agent installer wizard
- Checkbox adds --enable-commands flag to generated install commands
This allows users to:
1. Enable at install time via checkbox in the wizard
2. Toggle remotely via the Managed Agents UI for existing agents
3. Agents apply changes automatically on their next report cycle
- Add CommandsEnabled field to AgentInfo in pkg/agents/host/report.go
- Agent now reports whether AI command execution is enabled
- Server stores and exposes this via Host model
- Frontend can now show which agents have commands enabled
- This provides visibility before implementing remote configuration
Users who accumulated AI patrol findings before the patrol-without-AI
bug was fixed (24c4bb0b) could not dismiss them because the dismiss and
resolve endpoints required a Pro license.
Changes:
- Remove Pro license requirement from /api/ai/patrol/dismiss endpoint
- Remove Pro license requirement from /api/ai/patrol/resolve endpoint
- Add ClearAll() method to FindingsStore for bulk clearing
- Add DELETE /api/ai/patrol/findings endpoint for clearing all findings
- Add "Clear All" button to AI Insights UI
Users can now dismiss or resolve any findings they can see, and admins
can clear all findings at once if needed.
Backup and snapshot polling runs asynchronously and could take 20-45 seconds to complete, but WebSocket broadcasts happened on a separate fixed-interval timer. This caused frontend to show stale data until a broadcast happened to coincide with completed polling - which could take hours.
Now broadcasts state immediately after backup/snapshot polling completes, ensuring users see changes within seconds.
Related to #895
BREAKING CHANGE: AI command execution on agents is now disabled by default.
Users who want AI auto-fix must explicitly enable it with --enable-commands
flag or PULSE_ENABLE_COMMANDS=true environment variable.
Changes:
- Add --enable-commands flag (opt-in for command execution)
- Commands disabled by default for security (defense-in-depth)
- --disable-commands is now deprecated (logs warning, no longer needed)
- PULSE_DISABLE_COMMANDS deprecated in favor of PULSE_ENABLE_COMMANDS
- Update installer script to use --enable-commands
- Backwards compatibility: PULSE_DISABLE_COMMANDS=false still enables commands
This addresses community feedback about secure defaults for arbitrary
command execution on production infrastructure.
Related to #889
Extended Issue #891 fix to cover manual node addition via the UI:
1. HandleAddNode now checks for duplicates by Host URL (not name)
2. Disambiguator applied to PVE, PBS, and PMG node creation
3. Error message updated: 'host URL already exists' instead of 'name already exists'
This ensures the fix works whether nodes are added via:
- Agent auto-registration ✓
- Manual UI setup ✓
All node creation paths now consistently:
- Match by Host URL only
- Disambiguate duplicate hostnames with IP: 'px1' → 'px1 (10.0.2.224)'
Follow-up to #891 fix - also match by name+tokenID to handle the case
where the same physical host gets a new IP (DHCP). This ensures:
1. Same hostname + DIFFERENT token = different physical hosts → create separate nodes
2. Same hostname + SAME token = same host with new IP → update existing node
Also updates the host URL when an existing node is matched, so IP changes
are properly reflected in the saved configuration.
PROBLEM:
When two Proxmox hosts have the same hostname (e.g., 'px1' on different networks),
the auto-registration was matching by name and overwriting the first with the second.
This has been a recurring issue (#104) with at least 3 prior fix attempts.
ROOT CAUSE:
The auto-register handler matched existing nodes by BOTH Host URL and Name.
Matching by name is incorrect - different physical hosts can share hostnames.
FIXES:
1. Remove name-based matching in auto-registration - match by Host URL only
2. Add disambiguateNodeName() to append IP when duplicate hostnames exist
3. Add regression tests to prevent this from breaking again
Now when registering two hosts named 'px1':
- First becomes: px1
- Second becomes: px1 (10.0.2.224)
Both are stored as separate nodes with their own credentials.
Users who haven't enabled AI were seeing AI patrol findings from
heuristic analysis that they couldn't dismiss (license-gated).
- IsPatrolEnabled() now checks if Enabled is true
- IsAlertTriggeredAnalysisEnabled() also checks Enabled
- Updated tests to reflect new behavior
AI patrol and alert-triggered analysis require AI to be enabled
as a master switch. This prevents confusing UX where users see
AI features without having configured them.
Added request_timeout_seconds field to:
- AISettingsResponse struct (for GET responses)
- AISettingsUpdateRequest struct (for PUT requests)
- HandleUpdateAISettings handler logic (validation + persistence)
- HandleGetAISettings response builder
The frontend was already sending request_timeout_seconds but the
backend was ignoring it. Now the setting persists correctly.
Adds RequestTimeoutSeconds to AI config (default 300s / 5 min).
Users with low-power hardware running Ollama can increase this
value in Settings to prevent timeouts on slower inference.
Docker users were losing their license activation on every update because
/etc/machine-id changes when the container is recreated.
Changes:
- Store persistent encryption key in /data/.license-key (survives container updates)
- Fall back to machine-id for backwards compatibility with existing installations
- Existing users only need to re-enter their key once, then it persists forever
Fixes issue reported by customer upgrading from RC to v5.0.0.
When a PVE cluster has unique self-signed certificates on each node, Pulse
would mark secondary nodes as unhealthy because only the primary node's
fingerprint was used for all connections.
Now, during cluster discovery, Pulse captures each node's TLS fingerprint
and uses it when connecting to that specific node. This enables
"Trust On First Use" (TOFU) for clusters with unique per-node certs.
Changes:
- Add Fingerprint field to ClusterEndpoint config
- Add FetchFingerprint() to tlsutil for capturing node certs
- validateNodeAPI() now captures and returns fingerprints during discovery
- NewClusterClient() accepts endpointFingerprints map for per-node certs
- All client creation paths use per-endpoint fingerprints when available
Related to #879
rcourtman