Linux Fix Lab
Understand first.
Fix second.
Start with a symptom, a log line or a guided question. Separate the evidence, understand what a command tells you, and choose the next step with a way back.
Find a useful starting point
Search symptoms, commands and recognizable error messages. All problem guides below are also readable without JavaScript.
150 problems
No matching problems. Try a shorter term or a different category.
- Graphics, GPU & Display
AMDGPU ring timeout and GPU recovery
A stalled AMD command queue can freeze an application or the desktop; compare the first timeout, recovery result and software versions.
- Graphics, GPU & Display
AMDGPU cannot load a required firmware file
A missing AMD firmware blob can prevent acceleration or early display setup; check the exact requested filename and the initramfs contents.
- Graphics, GPU & Display
AMDGPU reports a GPU virtual-memory page fault
A GPU page fault identifies an invalid GPU address access; preserve the process and fault details before blaming physical RAM or video memory.
- Graphics, GPU & Display
NVIDIA Xid 79: GPU has fallen off the bus
NVIDIA Xid 79 indicates lost GPU accessibility over PCIe; preserve link and power evidence before repeated resets or a driver-only conclusion.
- Graphics, GPU & Display
NVIDIA driver and userspace library versions disagree
NVML or NVRM version mismatch often follows a driver update without reboot; compare the loaded module with installed userspace before reinstalling.
- Graphics, GPU & Display
NVIDIA module is rejected by signature enforcement
A signed-module policy can block NVIDIA after a kernel or DKMS update; check trust and build status before changing Secure Boot settings.
- Graphics, GPU & Display
Nouveau owns the GPU intended for NVIDIA
Only one kernel driver can own a GPU; confirm actual PCI binding before changing Nouveau blocklists or rebuilding an early boot image.
- Graphics, GPU & Display
NVIDIA suspend fails with video-memory preservation
NVIDIA video-memory preservation needs matching suspend integration and sufficient backing space; inspect parameters and units as one setup.
- Graphics, GPU & Display
Mesa falls back to CPU rendering
A desktop may remain usable through llvmpipe while GPU acceleration is missing; distinguish intended headless rendering from a failed driver path.
- Graphics, GPU & Display
Vulkan loader finds no usable driver
Vulkan initialization can fail before any GPU workload starts when ICD manifests, driver libraries or architecture-specific packages are missing.
- Graphics, GPU & Display
Application runs on the unintended GPU
Hybrid graphics separates display ownership from rendering; compare device selection per API before forcing the entire desktop onto another GPU.
- Graphics, GPU & Display
Monitor modes disappear because EDID is invalid
Bad or missing monitor identification can remove expected modes; separate cable, dock and KVM effects from an incorrect display configuration.
- Graphics, GPU & Display
DisplayPort link training fails after a modeset
A DisplayPort monitor can be detected yet fail link training; compare bandwidth demand and the physical path before changing GPU recovery knobs.
- Graphics, GPU & Display
Intel i915 reports a GPU hang
An i915 GPU hang may affect one engine or the whole desktop; preserve its error state and compare workload, kernel and userspace independently.
- Kernel & System Stability
Kernel Oops: NULL pointer dereference
A kernel NULL-pointer Oops is a kernel execution fault, not an ordinary application crash; preserve the first trace and loaded-module context.
- Kernel & System Stability
CPU soft lockup detected by the kernel watchdog
A soft lockup means the watchdog task did not run in time; inspect its CPU trace and distinguish a guest pause from a repeatable kernel stall.
- Kernel & System Stability
Kernel reports a task blocked for too long
Hung-task warnings identify prolonged uninterruptible waiting; inspect the wait path and preceding I/O events instead of killing the named task.
- Kernel & System Stability
RCU grace-period progress stalls
RCU stall warnings concern delayed kernel synchronization; compare CPU traces and timer progress before attributing the failure to RCU itself.
- Kernel & System Stability
Processor reports a machine-check hardware error
Machine-check records need CPU-family and bank-specific decoding; separate corrected reports from fatal events before replacing a component.
- Kernel & System Stability
PCIe AER reports corrected or uncorrected errors
PCIe AER records describe transaction or link errors with different severities; map the reporting port to its devices before blaming an endpoint.
- Kernel & System Stability
IOMMU blocks an invalid device DMA access
IOMMU fault records identify a device and DMA address domain; inspect its driver and passthrough setup while retaining isolation protections.
- Kernel & System Stability
Kernel module was built for a different kernel
A module’s version magic must fit the running kernel build; compare installed module metadata before rebuilding or selecting a boot entry.
- Kernel & System Stability
Kernel module cannot resolve a required symbol
Unknown-symbol errors can reflect missing module dependencies or ABI mismatch; inspect the exact symbol and dependency chain before forcing loads.
- Kernel & System Stability
A device callback aborts system suspend
Suspend can fail before entering a sleep state when a device callback rejects it; use the first PM error to separate rejection from wakeup.
- Kernel & System Stability
CPU hard lockup detected by a watchdog
A hard-lockup report means CPU interrupt heartbeats stopped; preserve watchdog traces and hardware context without assuming it is a GPU hang.
- Kernel & System Stability
CPU microcode update is absent or loads too late
CPU microcode must match the processor and early boot path; compare the active revision and update packaging without forcing a live reload.
- Boot, GRUB & systemd-boot
A required fstab mount sends boot into emergency mode
A missing non-root mount can block systemd’s local filesystem target; identify the exact fstab dependency before making optional mounts nonfatal.
- Boot, GRUB & systemd-boot
Initramfs cannot find the configured root UUID
An early shell with a missing root UUID can mean wrong identity, an unopened container or missing storage support; compare each layer in order.
- Boot, GRUB & systemd-boot
Encrypted root cannot be unlocked during boot
A LUKS boot prompt can fail from layout, missing key material or wrong device selection; keep existing keyslots intact while inspecting metadata.
- Boot, GRUB & systemd-boot
Installed bootloader and firmware boot mode disagree
A recovery image booted in legacy mode cannot inspect normal UEFI variables; confirm the installed boot path before reinstalling a bootloader.
- Boot, GRUB & systemd-boot
GRUB rescue cannot locate its normal module
GRUB rescue often means its prefix no longer points to readable modules; inspect GRUB’s own devices before attempting a persistent reinstall.
- Boot, GRUB & systemd-boot
systemd-boot does not list the installed Linux entry
systemd-boot reads entries from the ESP or XBOOTLDR; missing menu items require checking the actual boot partition, entry files and image paths.
- Boot, GRUB & systemd-boot
Secure Boot rejects a bootloader or kernel image
A boot-chain signature rejection occurs before normal driver loading; verify the signed loader, kernel and trust path without erasing firmware keys.
- Boot, GRUB & systemd-boot
Root filesystem check prevents normal boot
A boot-time fsck failure can require offline filesystem-specific repair; identify the checked device and exit status before bypassing checks.
- Boot, GRUB & systemd-boot
Boot waits for an obsolete hibernation resume device
A stale resume UUID can delay an otherwise valid cold boot after swap changes; identify resume configuration separately from root storage.
- Boot, GRUB & systemd-boot
LVM root logical volume is absent in the initramfs
LVM root requires visible physical volumes and early activation support; inspect metadata and initramfs contents before changing volume layout.
- Boot, GRUB & systemd-boot
Kernel cannot execute a working init process
No working init found means the kernel could not start userspace; inspect root selection, the init binary and its interpreter from rescue media.
- Boot, GRUB & systemd-boot
EFI boot partition is full during a kernel update
A full ESP or XBOOTLDR can leave a kernel update without a complete boot image; identify unused entries through the distribution’s cleanup tools.
- Storage & Filesystems
SATA CRC errors and repeated link resets
Separate SATA transport corruption from failing media when games or file copies stall alongside BadCRC or ICRC reports in the kernel log.
- Storage & Filesystems
ATA uncorrectable sectors and read errors
Treat repeatable UNC media errors as a data-preservation problem, using drive health and sector evidence before filesystem repair or stress tests.
- Storage & Filesystems
NVMe I/O timeouts and controller resets
Trace NVMe resets to controller health, PCIe evidence and kernel history instead of interpreting an I/O timeout as proof of one defective SSD.
- Storage & Filesystems
NVMe fails after idle or resume
Use a controlled NVMe power-state comparison when failures follow idle or resume; separate APST from PCIe ASPM and system suspend behavior.
- Storage & Filesystems
ext4 enters emergency read-only protection
Recognize ext4 protective write refusal after journal or metadata errors, including emergency_ro behavior that may coexist with visible rw mount flags.
- Storage & Filesystems
XFS shuts down after metadata or log I/O errors
Distinguish an XFS protective shutdown from a normal unmount, preserve the log, and choose recovery based on device stability and offline evidence.
- Storage & Filesystems
Btrfs ENOSPC despite apparent free space
Check Btrfs data and metadata allocation separately when snapshots or writes fail with ENOSPC although a simple free-space report looks sufficient.
- Storage & Filesystems
Btrfs checksum failures and unreadable extents
Interpret Btrfs checksum reports with scrub results and redundancy, then preserve damaged data without assuming scrub can repair a single copy.
- Storage & Filesystems
ext4 runs out of inodes before blocks
Identify ext4 inode exhaustion when creating tiny files fails while block space remains, and fix the file population rather than masking ENOSPC.
- Storage & Filesystems
Deleted files still occupy disk space
Explain df and du disagreement by locating deleted files still held open by processes, then release their handles without corrupting application state.
- Storage & Filesystems
A filesystem UUID in fstab no longer matches
Resolve a missing data mount after cloning or reformatting by comparing fstab identifiers with actual filesystem UUIDs and verifying device identity.
- Storage & Filesystems
NTFS3 refuses a dirty Windows volume
Respect NTFS dirty-volume protection and Windows hibernation state, using a clean shutdown and native filesystem checks before permitting Linux writes.
- Storage & Filesystems
USB storage resets while using UAS
Investigate USB mass-storage aborts by separating cable and power faults from bridge-specific UAS behavior, with a narrowly scoped fallback comparison.
- Storage & Filesystems
TRIM fails on a filesystem or storage layer
Check discard capability through the complete storage stack when fstrim reports unsupported operation, without confusing reported bytes with freed data.
- Audio, ALSA & PipeWire
ALSA cannot see the expected sound card
Locate a missing sound card below PipeWire by checking ALSA enumeration and kernel probing, before changing desktop routes or user audio settings.
- Audio, ALSA & PipeWire
PipeWire or its user session manager does not start
Separate a stopped user audio service from missing hardware, and inspect configuration or package errors before restarting the session repeatedly.
- Audio, ALSA & PipeWire
Sound goes to the wrong or muted output
Use PipeWire default-device and volume evidence to distinguish a muted sink or retained application route from a missing sound device or driver failure.
- Audio, ALSA & PipeWire
Another process holds the ALSA device
Identify exclusive ALSA access that blocks PipeWire or a direct client, then release the verified owner or route the application through the sound server.
- Audio, ALSA & PipeWire
PipeWire crackles or drops audio under load
Correlate audio glitches with PipeWire deadline counters and device timing, then compare buffer or workload changes one at a time with recoverable settings.
- Audio, ALSA & PipeWire
Bluetooth audio loses quality or its microphone
Distinguish Bluetooth playback and call profiles, check available roles and the active user session, and choose a profile suited to the intended task.
- Audio, ALSA & PipeWire
HDMI or DisplayPort audio has no usable sink
Check monitor audio capabilities and HDMI ELD presence before forcing an output profile, especially after docking, cable changes or display sleep.
- Audio, ALSA & PipeWire
S/PDIF stays silent or passthrough fails
Separate ordinary stereo PCM on S/PDIF from compressed passthrough, and check the digital profile, mixer switch and receiver capabilities before enabling codecs.
- Audio, ALSA & PipeWire
USB audio cannot set a sample rate or clock
Interpret USB-audio clock and sample-rate failures using advertised formats, interface controls and transport evidence rather than forcing a global rate.
- Audio, ALSA & PipeWire
WirePlumber 0.5 rejects an old 0.4 configuration
Migrate old WirePlumber Lua configuration to scoped SPA-JSON fragments after an upgrade, preserving device rules and avoiding outdated full-file overrides.
- Audio, ALSA & PipeWire
The microphone is visible but records silence
Trace microphone silence across PipeWire source selection, capture switches and application permissions, without mistaking a monitor source for a real mic.
- Audio, ALSA & PipeWire
Audio device access fails in the current session
Diagnose ALSA permission refusal and Bluetooth session ownership using login state and device ACL evidence, then repair access without broad device permissions.
- Networking, DNS & Wi-Fi
NetworkManager leaves an interface unmanaged
Identify who owns the interface when NetworkManager marks it unmanaged, then adjust the specific ownership rule without disrupting other networks.
- Networking, DNS & Wi-Fi
Wi-Fi blocked by rfkill or airplane mode
Separate software airplane-mode blocks from physical radio blocks before editing Wi-Fi profiles or replacing a driver that cannot enable the radio.
- Networking, DNS & Wi-Fi
Wi-Fi authentication loops or fails
Use handshake and EAP evidence to separate a personal Wi-Fi credential mismatch from enterprise authentication or driver errors before changing security.
- Networking, DNS & Wi-Fi
Wi-Fi driver cannot load compatible firmware
Distinguish an optional firmware version fallback from a final initialization failure, then match the adapter, kernel and distribution firmware package.
- Networking, DNS & Wi-Fi
Connected link but no DHCP IPv4 lease
Follow address acquisition after a working link or Wi-Fi authentication, separating a DHCP server or VLAN problem from DNS and duplicate-address refusal.
- Networking, DNS & Wi-Fi
Address connectivity works but DNS resolution fails
Inspect systemd-resolved’s upstream servers and resolv.conf integration when names fail, including DNS loops and validation errors without disabling safeguards.
- Networking, DNS & Wi-Fi
VPN works but internal names use the wrong DNS
Trace domain-specific DNS selection when VPN addresses work but internal hostnames fail, preserving public DNS and the intended private lookup path.
- Networking, DNS & Wi-Fi
Small requests work but large transfers stall
Investigate a path MTU mismatch when large transfers stall through VPNs or tunnels, using destination-specific routing and measured packet-size evidence.
- Networking, DNS & Wi-Fi
Ethernet has no carrier or repeatedly loses link
Check physical Ethernet carrier and negotiated link mode before DHCP or DNS changes when a cable connection never becomes operational or keeps dropping.
- Networking, DNS & Wi-Fi
Network transmit queue watchdog or hardware hang
Investigate driver-level transmit stalls when the kernel watchdog fires, using the exact NIC and driver version before testing a narrowly scoped workaround.
- Networking, DNS & Wi-Fi
IPv6 has an address but no usable default route
Separate IPv6 addressing from router-advertisement route learning when IPv4 works but IPv6 lacks a gateway, especially on hosts with forwarding enabled.
- Networking, DNS & Wi-Fi
IPv4 conflict or IPv6 duplicate address detection
Resolve confirmed address conflicts when activation is refused or an IPv6 address remains unusable, retaining conflict detection instead of bypassing it.
- Networking, DNS & Wi-Fi
VPN and container routes overlap a local network
Find the actual route chosen for an affected address when VPN, VM or container subnets overlap, then correct the specific collision with a recovery path.
- Networking, DNS & Wi-Fi
Desktop says limited connectivity despite working access
Interpret NetworkManager’s optional connectivity probe when a desktop warning disagrees with actual access, including portal, filtering and route-metric effects.
- SSH, Security & Permissions
SSH ignores an authorized public key
A valid public key can be ignored when its file or parent directories are writable by other users. Inspect the server-side path first.
- SSH, Security & Permissions
SSH reports a changed server host key
A host-key warning can follow a rebuild or a wrong destination. Verify the new fingerprint independently before editing known_hosts.
- SSH, Security & Permissions
SSH exhausts authentication attempts before the right key
An agent with many keys can use up the server’s authentication limit. Check identity selection before increasing server-wide limits.
- SSH, Security & Permissions
sshd rejects a configuration change
An unknown option or invalid Match placement can block an SSH reload. Validate configuration and retain a working recovery session.
- SSH, Security & Permissions
SSH peers cannot agree on algorithms
Older appliances may offer algorithms disabled by a current client. Distinguish host-key and key-exchange negotiation before acting.
- SSH, Security & Permissions
sudo rejects an account or command
Correct authentication does not grant a command absent from sudo policy. Compare current group membership and the exact allowed command.
- SSH, Security & Permissions
sudo has no channel for password input
A remote script or GUI launcher can lack a terminal for sudo authentication. Select an appropriate prompt without embedding passwords.
- SSH, Security & Permissions
An ACL mask restricts access despite an explicit grant
A named ACL entry can show write permission while its mask removes effective write access. Inspect effective rights and parent traversal.
- SSH, Security & Permissions
An executable is stored on a noexec mount
Executable permission bits do not override a noexec mount. Inspect the mount governing the actual path before changing file permissions.
- SSH, Security & Permissions
AppArmor blocks an application operation
An AppArmor denial names the profile and operation that were blocked. Check the intended path before adding a narrowly reviewed rule.
- SSH, Security & Permissions
SELinux denies a service operation
An enforcing AVC denial identifies process and target contexts. Compare them with intended policy before considering any local allow rule.
- SSH, Security & Permissions
A file’s SELinux label differs from its path policy
Moved files can retain an old SELinux label. Compare the expected path context and preview relabeling before altering security policy.
- Desktop, X11 & Wayland
An X11 application cannot open its display
A stale DISPLAY or missing X authorization can break one launch context. Compare the failing process with the working desktop session.
- Desktop, X11 & Wayland
Qt cannot initialize its platform plugin
A found Qt platform plugin can still fail to load due to dependencies or mixed versions. Inspect plugin selection before reinstalling everything.
- Desktop, X11 & Wayland
Xorg exits with no screens found
The final no-screens message summarizes an Xorg startup failure. Read the earlier device and configuration errors to identify the blocked step.
- Desktop, X11 & Wayland
The session runtime directory is missing or invalid
Wayland and desktop IPC depend on a private runtime directory. Check ownership, mode and login lifetime before changing session variables.
- Desktop, X11 & Wayland
A desktop portal cannot activate its backend
A missing file picker or screen-sharing dialog can reflect wrong portal selection. Inspect backend support and the desktop session environment.
- Desktop, X11 & Wayland
A desktop process cannot reach its session D-Bus
Applications launched outside the desktop can inherit a missing or obsolete session bus address. Compare process context before starting another bus.
- Desktop, X11 & Wayland
Plasma global shortcuts stop responding
One broken shortcut can be a conflicting assignment; many broken shortcuts may reflect session integration. Inspect the affected component first.
- Desktop, X11 & Wayland
A GNOME Shell extension breaks after a Shell update
GNOME Shell API changes can make an extension incompatible. Compare supported Shell versions and disable the affected extension in isolation.
- Desktop, X11 & Wayland
Plasma applications have unexpected display scaling
Different output scale settings and Qt overrides can multiply application scaling. Compare native output settings with the app’s launch environment.
- Desktop, X11 & Wayland
XKB cannot compile the configured keyboard layout
A layout typo or missing include can prevent keymap compilation. Test the selected names without applying an experimental map to the session.
- Desktop, X11 & Wayland
Files or links open in the wrong default application
MIME defaults and desktop-specific overrides can select an unexpected application. Compare detected type, handler ID and configuration precedence.
- Desktop, X11 & Wayland
A privileged desktop action has no authentication agent
A working polkit daemon does not guarantee a password dialog in a custom desktop. Inspect the session and its registered authentication agent.
- Package Management & Updates
Another package manager holds the APT/dpkg lock
APT lock contention often means an update is still running. Identify the holder and its progress; deleting dpkg lock files is not a repair.
- Package Management & Updates
dpkg has pending configuration after an interrupted transaction
An interrupted install can leave packages unpacked but unconfigured. Audit the recorded states before resuming scripts or removing packages.
- Package Management & Updates
APT cannot resolve the requested dependencies
Held packages, pins and mixed release sources can make a dependency request impossible. Compare candidate versions and the simulated transaction.
- Package Management & Updates
APT rejects a repository’s signature or trust setup
Missing keys or an unsigned archive can block APT metadata updates. Verify the repository’s signing key and scoped Signed-By configuration.
- Package Management & Updates
APT cannot find a repository Release file
A wrong suite, obsolete release or incorrect repository URI can remove the Release path APT expects. Check the source against the installed OS.
- Package Management & Updates
pacman cannot acquire its database lock
A running libalpm writer or a stale db.lck after an interruption can block pacman. Prove which case applies before touching the lock file.
- Package Management & Updates
pacman refuses to overwrite a conflicting file
A conflicting path may belong to another package or a manual installation. Query ownership before removing files or using overwrite options.
- Package Management & Updates
A partial Arch upgrade leaves incompatible libraries
Updating only part of a rolling system can break library consumers or pacman itself. Compare transaction history and ELF dependencies without symlink hacks.
- Package Management & Updates
pacman cannot trust or validate a package signature
Keyring age, clock errors and damaged downloads can break signature checks. Verify the signer and trust chain before refreshing any key material.
- Package Management & Updates
DNF cannot satisfy a package dependency plan
A missing provider or incompatible repository package can block DNF. Compare installed requirements and cached providers before allowing removals.
- Package Management & Updates
A Flatpak app’s required runtime is unavailable
A Flatpak runtime must match its ID, architecture, branch and installation scope. Compare app metadata, installed refs and configured remotes.
- Package Management & Updates
A Flatpak app cannot see a host file
Host filesystem access and portal grants are separate from Unix file modes. Inspect app permissions before enlarging the sandbox’s access.
- Services & systemd
Service stops before its executable starts
A service exits with 203/EXEC although a similar shell command works. Check the exact executable, interpreter, mount flags and unit context.
- Services & systemd
WorkingDirectory prevents a service from starting
A unit can fail with 200/CHDIR before launching its program. Resolve missing mounts, traversal permissions and the directory visible inside its root.
- Services & systemd
Service user or group setup fails
A service fails with USER or GROUP before its program runs. Separate absent accounts, NSS lookup problems and credential or user namespace restrictions.
- Services & systemd
A notify service never becomes ready
The process exists but Type=notify remains activating until its deadline. Check readiness messages, sender attribution and blocked initialization work.
- Services & systemd
Crash loops reach the systemd start limit
Repeated restarts eventually stop with start-limit-hit. Find the earliest application failure before changing backoff or resetting the start counter.
- Services & systemd
A failed prerequisite prevents service activation
A service ends with a dependency result without running its own program. Trace required jobs and ordering before weakening dependencies or adding delays.
- Services & systemd
Service sandbox setup fails with NAMESPACE
A unit fails with 226/NAMESPACE before execution. Inspect sandbox paths, root images and host capabilities to identify the specific isolation failure.
- Services & systemd
Managed state or runtime directories cannot be prepared
StateDirectory or RuntimeDirectory fails before ExecStart. Check path collisions, ownership and capacity while preserving existing application state.
- Services & systemd
A timer runs late or misses offline periods
Scheduled jobs miss the expected time. Check calendar interpretation, activation, random delays and whether offline calendar runs should be caught up.
- Services & systemd
Socket activation cannot bind its listening address
A .socket unit fails because another listener owns its endpoint. Identify the owner and intended activation model before changing bindings or stopping it.
- Services & systemd
Journald suppresses a burst of service messages
A noisy service has log gaps with suppression notices. Distinguish journal rate limits from application-side logging and storage retention settings.
- Services & systemd
A service misses its userspace watchdog heartbeat
A running service aborts after missing WATCHDOG=1 heartbeats. Check notification support, sender identity and workload stalls before changing its deadline.
- NixOS & Configuration
A configuration option does not exist
Evaluation rejects an option before any system build. Check spelling, imported modules and the exact Nixpkgs revision instead of guessing replacement names.
- NixOS & Configuration
Two modules define incompatible option values
Evaluation cannot merge a scalar option defined in several modules. Trace definition locations and priorities before applying a force override.
- NixOS & Configuration
Module evaluation enters an infinite recursion
Nix cannot resolve a configuration dependency cycle. Inspect config-dependent imports, self-referencing values and argument timing before restructuring modules.
- NixOS & Configuration
A flake cannot see a newly created source file
A file exists in the checkout but is absent from the copied flake source. Check Git tracking, ignored paths and the import relative to the flake root.
- NixOS & Configuration
Pure evaluation rejects a host-local file
A flake reads a path outside its declared inputs and pure evaluation rejects it. Distinguish build-time source data from runtime filesystem paths.
- NixOS & Configuration
The Nix command requires a disabled experimental feature
Nix rejects flakes or the modern CLI before evaluation. Check the active binary and settings scope, then enable only the features the workflow needs.
- NixOS & Configuration
Nix rejects a store path's cache signature
A substitute or copied store path lacks a trusted signature. Verify the cache identity, public key and daemon configuration before changing trust rules.
- NixOS & Configuration
A fetched fixed-output source has a different hash
A fetcher output differs from its expected hash. Verify the pinned revision, artifact and hashing mode before accepting the newly reported digest.
- NixOS & Configuration
An unprivileged user cannot reach the Nix daemon
Nix operations fail at the daemon socket or admission policy. Separate socket access, allowed-users and administrator-level trusted-users privileges.
- NixOS & Configuration
Garbage collection frees less space than expected
The Nix store remains large after collection because roots retain closures. Inspect generations and result links before deleting recovery paths or store files.
- NixOS & Configuration
A tested generation disappears after reboot
Runtime changes worked but reboot restores an older configuration. Compare running, booted and profile targets and choose the intended rebuild action.
- NixOS & Configuration
A generic Linux binary cannot find its loader on NixOS
An existing executable fails before normal startup because it expects an FHS loader or libraries. Inspect ELF metadata and choose a scoped packaging route.
- Performance & Virtualization
A service is slow while the host still has idle CPU
CPU quota can suspend a busy cgroup while other CPUs remain idle. Compare quota settings and throttling deltas before adding threads or changing governors.
- Performance & Virtualization
CPU frequency remains low under sustained work
Low reported frequency can reflect policy caps, thermal limits or a requested frequency rather than actual clock speed. Compare policy and sensor evidence.
- Performance & Virtualization
Memory reclaim stalls interactive work
Applications pause during memory pressure before any OOM kill. Use memory PSI, available memory and reclaim deltas to separate pressure from a CPU bottleneck.
- Performance & Virtualization
A service hits its cgroup memory boundary
A process is killed although host RAM remains available. Correlate cgroup limits, memory.events and the kill record before increasing a service budget.
- Performance & Virtualization
Swap activity causes repeated interactive pauses
Occupied swap is not itself a fault. Correlate swap-in/out rates, memory pressure and working-set demand before changing swappiness or disabling swap.
- Performance & Virtualization
Background writeback stalls foreground work
Foreground latency rises during bulk writes without device errors. Compare I/O PSI, queue latency and dirty pages before changing writeback or I/O budgets.
- Performance & Virtualization
A service cannot create more threads or processes
Fork or thread creation fails while memory remains available. Check cgroup task counts and ancestor limits before raising TasksMax or blaming the application.
- Performance & Virtualization
Perf is denied access to performance events
Perf cannot enable an event under the current security scope. Check paranoid policy, capabilities and container restrictions before enabling broad monitoring.
- Performance & Virtualization
Hardware counters are unavailable to perf
Perf software events work while CPU cycles or instructions are unavailable. Check PMU exposure, supported event names and the guest CPU model.
- Performance & Virtualization
CPU placement and memory policy do not align
A NUMA workload runs far from its memory or fails policy setup. Inspect topology, allowed nodes and process page placement before adding hard bindings.
- Performance & Virtualization
QEMU cannot initialize KVM acceleration
A guest fails to start with KVM or runs using a slower software accelerator. Check hardware support, /dev/kvm access and the chosen domain type.
- Performance & Virtualization
A guest loses CPU time to host contention or quotas
A guest feels CPU-starved despite unused virtual CPUs. Compare guest steal time with host placement and quotas before adding more vCPUs.
Read the signals in your log
163 explainable signatures for kernel, systemd, graphics, audio, storage and more. Multiple findings remain separate; this tool does not assign a universal diagnosis.
Up to 2 million characters / 20,000 lines. Lines above 4,096 characters are truncated with a notice. Logs are not stored; changing page or language clears the excerpt. Commands below are read-only recommendations, never automatically executed.
Automatic redaction may miss sensitive data. Review excerpts before giving a file or clipboard contents to another person or service. No export is submitted by this page.
One question at a time
Choose a topic. Each assistant explains the evidence to collect, distinguishes failure paths and links to the relevant guides. Answers stay in this tab; only the current question is kept in the address for language switching.
Evidence, with limits clearly stated
The guides reference primary project and distribution sources. Review dates mean editorial source checks; they are not hardware test results. Synthetic logs test the detector, not a promised cure. Driver, kernel and distribution behavior can differ between installed releases.
Learn to capture and read journalctl output · Browse the practical Linux guides