Feature: NVML encode/decode % included in nvidia GPU usage metric#2
Merged
Conversation
…of-three Query NVML encoder_utilization() and decoder_utilization() alongside existing utilization_rates() for GPU compute. Combine all three into a single composite percentage using max()-of-three approach, ensuring the device appears busy when any engine (compute, encode, decode) is active. This prevents premature sleep inhibition release during video encoding/ decoding workloads that would otherwise show zero GPU compute usage.
NVML utilization_rates() reports SM busy percentage at the current clock speed, not normalized to max rated frequency. A GPU running 200MHz at 100% is effectively ~6% loaded vs its 3200MHz peak — the same principle as CPU freq-weighting (cpu.rs:404-436). Apply identical formula for NVIDIA GPUs: effective_max = max(current_freq, rated_max, observed_peak) weighted_usage = raw_gpu_pct * (current_freq / effective_max) Track per-card peak observed frequency to handle turbo boost beyond rated maximums. Encoder/decoder engines run at fixed clocks and are combined via max() without freq weighting. AMD/Intel remain unchanged (their sysfs gpu_busy_percent is already normalized by the kernel).
…nc config defaults - Remove redundant debug lines from cpu.rs (core count), gpu.rs (NVML init, enumeration count, per-card lines) - Rewrite network.rs collect() to return NetworkThroughput with per-interface Mbps breakdown; remove 'Skipping excluded interface' debug line - Update disk.rs collect() to return DiskThroughput with interval + total + per-device MB/s breakdown using delta tracking in loop - Add GpuDisplayEntry, sorted_gpu_display(), gpu/display_string(), network/ disk display formatters in metrics/mod.rs - Create TickMetrics struct carrying full throughput alongside Metrics - Refactor service.rs tick() to use new types; GPU log shows sorted entries with inline smoothed values per-GPU; Network/Disk show breakdowns - Sync config defaults: total_threshold=25.0, gpu threshold=15.0, network/ disk thresholds=10.0 across src/config.rs and all docs
… in display HashMap iteration order is non-deterministic, causing log output to show different ordering on every tick. Sort key-value pairs by name before formatting so each run produces identical alphabetical order.
- Implement 3-tier config merging (embedded -> /etc/rouser -> $XDG_CONFIG_HOME) - Auto-install default config on first startup via load_merged() - --print-config serializes merged TOML instead of raw embedded string - Fix tracing init order so auto-install INFO log is captured at startup - Remove which crate dependency and NOTIFY_SOCKET warning (not used) - Update install.sh to respect $XDG_BIN_HOME, $XDG_CONFIG_HOME - Remove config file copying from install script (app creates it at startup) - Remove gh CLI dependency from install script (curl works fine) - Fix systemd service path: extracted/systemd/rouser.service - Correct maintainer name: Owain Jones in deb.sh, rpm.sh, pkgbuild.sh - Add XDG Base Directory Compliance section to AGENTS.md - Add vibecoded development note to README
owaindjones
added a commit
that referenced
this pull request
May 2, 2026
Bug #1: 'Predictive cooldown extension' info log fired on every tick while extended cooldown was active because predicted_additional_time was already set from a previous tick. Added check for predicted_additional_time.is_zero() so the message only logs once per transition into below-threshold state, matching how 'Sleep inhibited' logs only fire on state transitions. Bug #2: Predictive cooldown extension had no effect — inhibition was released after base cooldown_duration (10s) instead of respecting the predicted +1028s extension. The release logic checked plain cooldown_duration first and released before reaching the predictive branch. Replaced two-branch logic with single path using std::cmp::max(cooldown_duration, predicted_additional_time) so the prediction always extends (not replaces) the base cooldown period.
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
No description provided.