Root cause: sda's temperature exceeded threshold in the past, causing
smartctl to return exit code 32 (warning: "Attributes have been <= threshold
in the past"). The agent checked output.status.success() and rejected the
entire output as failed, even though the data (serial, temperature, health)
was perfectly valid.
Smartctl exit codes are bit flags for informational warnings:
- Exit 0: No warnings
- Exit 32 (bit 5): Attributes were at/below threshold in past
- Exit 64 (bit 6): Error log has entries
- etc.
The output data is valid regardless of these warning flags.
Solution: Parse output as long as it's not empty, ignore exit code.
Only return UNKNOWN if output is actually empty (command truly failed).
Result: Data_3 will now show "ZDZ4VE0B T: 31°C" instead of "? Data_3: sda"
Bump version to v0.1.225
Root cause: SMART data was collected TWICE:
1. Sequential collection during pool detection in get_drive_info_for_path()
using problematic tokio::task::block_in_place() nesting
2. Parallel collection in get_smart_data_for_drives() (v0.1.223)
The sequential collection happened FIRST during pool detection, causing
sda (Data_3) to timeout due to:
- Bad async nesting: block_in_place() wrapping block_on()
- Sequential execution causing runtime issues
- sda being third in sequence, runtime degraded by then
Solution: Remove SMART collection from get_drive_info_for_path().
Pool drive temperatures are populated later from the parallel SMART
collection which properly uses futures::join_all.
Benefits:
- Eliminates problematic async nesting
- All SMART queries happen once in parallel only
- sda/Data_3 should now show serial (ZDZ4VE0B) and temperature
Bump version to v0.1.224
Root cause: SMART data was collected sequentially, one drive at a time.
With 5 drives taking ~500ms each, total collection time was 2.5+ seconds.
When disk collector runs every 1 second, this caused overlapping
collections creating resource contention. The last drive (sda/Data_3)
would timeout due to the drive being accessed by the previous collection.
Solution: Query all drives in parallel using futures::join_all. Now all
drives get their SMART data collected simultaneously with independent
3-second timeouts, eliminating contention and reducing total collection
time from 2.5+ seconds to ~500ms (the slowest single drive).
Benefits:
- All drives complete in ~500ms instead of 2.5+ seconds
- No overlapping collections causing resource contention
- Each drive gets full 3-second timeout window
- sda/Data_3 should now show temperature and serial number
Bump version to v0.1.223
Root cause: run_command_with_timeout() was calling cmd.spawn() without
configuring stdout/stderr pipes. This caused command output to go to
journald instead of being captured by wait_with_output(). The disk
collector received empty output and failed silently.
Solution: Configure stdout(Stdio::piped()) and stderr(Stdio::piped())
before spawning commands. This ensures wait_with_output() can properly
capture command output.
Fixes: Empty Storage section, lsblk output appearing in journald
Bump version to v0.1.222
v0.1.220 broke disk collector by changing the import from
std::process::Command to tokio::process::Command, but lines 193 and
767 explicitly used std::process::Command::new() which silently failed.
Solution: Import both as aliases (TokioCommand/StdCommand) and use
appropriate type for each operation - async commands use TokioCommand
with run_command_with_timeout, sync commands use StdCommand with
system timeout wrapper.
Fixes: Empty Storage section after v0.1.220 deployment
Bump version to v0.1.221
- Changed disk collector to use tokio::process::Command instead of std::process::Command
- Updated run_command_with_timeout to properly kill processes on timeout
- Fixes issue where smartctl hangs on problematic drives (/dev/sda) freezing entire agent
- Timeout now force-kills hung processes using kill -9, preventing orphaned smartctl processes
This resolves the issue where Data_3 showed unknown status because smartctl was hanging
indefinitely trying to read from a problematic drive, blocking the entire collector.
Bump version to v0.1.220
Co-Authored-By: Claude <noreply@anthropic.com>
- Update BackupData structure to support multiple backup disks
- Scan /var/lib/backup/status/ directory for all status files
- Calculate status icons for backup and disk usage
- Aggregate repository status from all disks
- Update dashboard to display all backup disks with per-disk status
- Display repository list with count and aggregated status
- Agent now extracts "C" + digits pattern (C3, C10) using char parsing
- Removes suffixes like "_ACPI", "_MWAIT" at source
- Reduces JSON payload size over ZMQ
- No regex dependency - uses fast char iteration (~1μs overhead)
- Robust fallback to original name if pattern not found
- Dashboard simplified to use clean names directly
Bump version to v0.1.212
Co-Authored-By: Claude <noreply@anthropic.com>
- Changed code to use zmq.transmission_interval_seconds instead of top-level collection_interval_seconds
- Removed collection_interval_seconds from AgentConfig
- Updated validation to check zmq.transmission_interval_seconds
- Improves config organization by grouping all ZMQ settings together
Bump version to v0.1.210
Co-Authored-By: Claude <noreply@anthropic.com>
- Changed CpuData.cstate from String to Vec<CStateInfo>
- Added CStateInfo struct with name and percent fields
- Collector calculates percentage for each C-state based on accumulated time
- Sorts and returns top 3 C-states by usage
- Dashboard displays: "C10:79% C8:10% C6:8%"
Provides better visibility into CPU idle state distribution.
Bump version to v0.1.209
Co-Authored-By: Claude <noreply@anthropic.com>
- Changed CpuData.frequency_mhz to CpuData.cstate (String)
- Implemented collect_cstate() to read CPU idle depth from sysfs
- Finds deepest C-state with most accumulated time (C0-C10)
- Updated dashboard to display C-state instead of frequency
- More accurate indicator of CPU activity vs power management
Bump version to v0.1.208
Co-Authored-By: Claude <noreply@anthropic.com>
The service_status_cache from discovery only has active_state with
all detailed metrics set to None. During collection, get_service_status()
was returning cached data instead of fetching fresh systemctl show data.
Now always fetch fresh data to populate memory_bytes, restart_count,
and uptime_seconds properly.
Shared:
- Add memory_bytes, restart_count, uptime_seconds to ServiceData
Agent:
- Add new fields to ServiceStatusInfo struct
- Fetch MemoryCurrent, NRestarts, ExecMainStartTimestamp from systemctl show
- Calculate uptime from start timestamp
- Parse and populate new fields in ServiceData
- Remove unused load_state and sub_state fields
Dashboard:
- Add memory_bytes, restart_count, uptime_seconds to ServiceInfo
- Update header: Service, Status, RAM, Uptime, ↻ (restarts)
- Format memory as MB/GB
- Format uptime as Xd Xh, Xh Xm, or Xm
- Show restart count with ! prefix if > 0 to indicate instability
All metrics obtained from single systemctl show call - zero overhead.
Complete removal of service resource metrics:
Agent:
- Remove memory_mb and disk_gb fields from ServiceData struct
- Remove get_service_memory_usage() method
- Remove get_service_disk_usage() method
- Remove get_directory_size() method
- Remove unused warn import
Dashboard:
- Remove memory_mb and disk_gb from ServiceInfo struct
- Remove memory/disk display from format_parent_service_line
- Remove memory/disk parsing in legacy metric path
- Remove unused format_disk_size() function
Service resource metrics were slow, unreliable, and never worked
properly since structured data migration. Will be handled differently
in the future.
Service control migrated to SSH, command receiver no longer needed.
- Remove command_receiver Socket from ZmqHandler
- Remove try_receive_command method
- Remove AgentCommand enum
- Remove command_port from ZmqConfig
Changes:
- Rename docker images from 'image_node:18...' to 'I node:18...' for conciseness
- Change image status from 'active' to 'inactive' for neutral informational display
- Images now show with gray empty circle ○ instead of green filled circle ●
Docker images are static artifacts without meaningful operational status, so using inactive status provides neutral gray display that won't trigger alerts or affect service status aggregation.
Fixes random host disconnections caused by blocking operations preventing timely ZMQ packet transmission.
Changes:
- Add run_command_with_timeout() wrapper using tokio for async command execution
- Apply 10s timeout to smartctl (prevents 30+ second hangs on failing drives)
- Apply 5s timeout to du, lsblk, systemctl list commands
- Apply 3s timeout to systemctl show/is-active, df, ip commands
- Apply 2s timeout to hostname command
- Use system 'timeout' command for sync operations where async not needed
Critical fixes:
- smartctl: Failing drives could block for 30+ seconds per drive
- du: Large directories (Docker, PostgreSQL) could block 10-30+ seconds
- systemctl/docker: Commands could block indefinitely during system issues
With 1-second collection interval and 10-second heartbeat timeout, any blocking operation >10s causes false "host offline" alerts. These timeouts ensure collection completes quickly even during system degradation.
Agent changes:
- Changed docker ps and docker images commands to run without sudo
- cm-agent user is already in docker group, so sudo is not needed
- Fixes "unable to change to root gid: Operation not permitted" error
- Systemd security restrictions were blocking sudo gid changes
This fixes Docker container and image collection on systems with
systemd security hardening enabled.
Updated to version 0.1.178
Agent changes:
- Log stderr output when docker images command fails
- This will show the actual error message (e.g., permission denied, docker not found)
- Helps diagnose why docker images collection is failing
Updated to version 0.1.177
Agent changes:
- Changed debug!() to info!() for Docker collection logs
- This allows logs to show with default RUST_LOG=info setting
- Added info import to tracing use statement
Now logs will be visible in journalctl without needing to change log level:
- "Collecting Docker sub-services for service: docker"
- "Found X Docker containers"
- "Found X Docker images"
- "Total Docker sub-services added: X"
Updated to version 0.1.176
Agent changes:
- Added debug logging to Docker images collection function
- Log when Docker sub-services are being collected for a service
- Log count of containers and images found
- Log total sub-services added
- Show command failure details instead of silently returning empty vec
This will help diagnose why Docker images aren't showing up as sub-services
on some hosts. The logs will show if the docker commands are failing or if
the collection is working but data isn't being transmitted properly.
Updated to version 0.1.175
Agent changes:
- Added get_docker_images() function to list all Docker images
- Use docker images to show stored images with repository:tag and size
- Display images as sub-services under docker service with size in parentheses
- Skip dangling images (<none>:<none>)
- Images shown with active status (always present when listed)
Example display:
● docker active 139M 1MB
├─ ● docker_gitea active
├─ ○ docker_old-app inactive
├─ ● image_nginx:latest (142MB)
├─ ● image_postgres:15 (379MB)
└─ ● image_gitea:latest (256MB)
Updated to version 0.1.174
Agent changes:
- Use docker ps -a to show ALL containers (running and stopped)
- Map container status: Up -> active, Exited/Created -> inactive, other -> failed
- Display Docker containers as sub-services under the docker service
- Each container shown with proper status indicator
Example display:
● docker active 139M 1MB
├─ ● docker_gitea active
├─ ○ docker_old-app inactive
└─ ● docker_immich active
Updated to version 0.1.173
Agent changes:
- Changed docker ps to docker ps -a to show ALL containers (running and stopped)
- Map container status: Up -> active, Exited/Created -> inactive, other -> failed
- Display Docker containers as individual top-level services instead of sub-services
- Each container shown as "docker_{container_name}" in service list
This provides better visibility of all containers and their status directly in the
services panel, making it easier to see stopped containers at a glance.
Updated to version 0.1.172
Agent changes:
- Parse /proc/net/vlan/config to extract VLAN IDs for interfaces
- Detect primary physical interface via default route
- Auto-assign primary interface as parent for virtual interfaces without explicit parent
- Added vlan_id field to NetworkInterfaceData
Dashboard changes:
- Display VLAN ID in format "interface (vlan X): IP"
- Show VLAN IDs for both nested and standalone virtual interfaces
This ensures virtual interfaces (docker0, tailscale0, etc.) are properly nested
under the primary physical NIC, and VLAN interfaces show their IDs.
Updated to version 0.1.170
Agent changes:
- Filter out ifb* interfaces from network display
- Parse @parent notation for VLAN interfaces (e.g., lan@enp0s31f6)
- Show physical interfaces even without IP addresses
- Only filter virtual interfaces that have no IPs
- Extract parent interface relationships for proper nesting
Dashboard changes:
- Nest VLAN/child interfaces under their physical parent
- Show physical NICs with status icons even when down
- Display child interfaces grouped under parent interface
- Keep standalone virtual interfaces at root level
Updated to version 0.1.169
Move network collection from NixOS collector to dedicated NetworkCollector. Add link status detection for physical interfaces (up/down). Group interfaces by physical/virtual, show status icons for physical NICs only. Down interfaces show as Inactive instead of Critical.
Version bump to 0.1.165
Extend NixOS collector to gather network interfaces using ip command JSON output. Display all interfaces with IPv4 and IPv6 addresses in Network section above CPU metrics. Filters out loopback and link-local addresses.
Version bump to 0.1.161
Update systemd collector to use sudo for docker ps command to resolve
permission issues when cm-agent user lacks docker group membership.
This ensures Docker containers are properly discovered and displayed
as sub-services under the docker service.
Version: 0.1.160
- Support Temperature_Case attribute for Intel SSDs
- Support Media_Wearout_Indicator attribute for wear percentage
- Parse wear value from column 3 (VALUE) for Media_Wearout_Indicator
- Fixes temperature and wear display for Intel PHLA847000FL512DGN drives
- Fix physical drive serial number display in dashboard
- Improve pool health calculation for arrays with multiple disks
- Support proper tree symbols for multiple parity drives
- Read git commit hash from /var/lib/cm-dashboard/git-commit for Build display
- Fix NVMe serial number parsing to handle whitespace variations
- Move mount point to MergerFS header, remove drive count
- Restructure data drives to same level as parity with Data_1, Data_2 labels
- Remove "Total:" label from pool usage line
- Update parity to use closing tree symbol as last item
- Add serial_number field to DriveData structure
- Collect serial numbers from SMART data for all drives
- Display truncated serial numbers (last 8 chars) in dashboard
- Fix parity drive label to show status icon before "Parity:"
- Fix mount point label styling to match other labels
Updates disk collector and dashboard to show drive serial numbers
instead of device names (sdX) for MergerFS data/parity drives.
Agent extracts serial numbers from SMART data and dashboard
displays them when available, falling back to device names.
