NVSentinel detects and remediates GPU faults on Kubernetes nodes
C-Rank Grade: A (Welcoming) - 66/100
External PR Merge Rate: 73%
Response Time: 3d
First Timer Success: 58%
NVIDIA/NVSentinel has a recorded first-timer success rate of 57.7%. Repositories ranked A typically provide actionable feedback during code reviews and actively nurture new community contributors.
Maintainers in NVIDIA/NVSentinel respond to incoming external pull requests in approximately 70.1 hours on average. Keeping PRs focused on single tasks and ensuring tests pass helps maintainers review faster.
The C-Rank™ system evaluates GitHub projects on a 0–100 scale using real data: PR merge rates, review turnaround time, active maintainer presence, and first-time contributor success. A score of 65.5 places NVIDIA/NVSentinel in the A tier.
The external contributor pull request merge rate for NVIDIA/NVSentinel is 73.5%, based on public PR activity from non-core contributors.
NVIDIA/NVSentinel currently has 3 active issue(s) tagged with beginner-friendly labels like "good first issue", "beginner", or "up-for-grabs".
Highly welcoming maintainers in NVIDIA/NVSentinel. Prompt code reviews with positive guidance for new contributors.
Average Response Latency
Tracks hours until a maintainer leaves a review, comment, or PR response.
External Acceptance Rate
Percentage of community pull requests successfully merged into main.
First PR Conversion
Rate at which developers submitting their first repository PR succeed.
Show contributors that your repository actively reviews and merges external pull requests.
[](https://getmerged.abhishekco.de/NVIDIA/NVSentinel?utm_source=github&utm_medium=badge)
Historical health & responsiveness evolution
External PR acceptance
First PR success rate
50% receive response in
90% receive response in
Maintainer review activity heatmap & prime window prediction
Calculating local time...
Calculating local time...
Detected: UTC
Maintainers are most responsive around 06:00 UTC (06:00 UTC in your timezone). Recommended submission window is Tuesday - Thursday from 06:00 to 10:00 UTC (06:00 - 10:00 UTC local).
Maintainer Strategy: Maintainer Sweet Spot: PRs under 150 lines merge in ~46 hours (92% approval). Large PRs (>500 lines) take ~336 hours (33% approval). Keep your diff compact for rapid review.
Prerequisites I searched existing issues I can reproduce this issue Code of Conduct I agree to follow NVSentinel's Code of Conduct Bug Description It's possible for the kubernetes-object-monitor to not send a healthy event for a deleted object if it misses the delete event fired from the object's corresponding informer. Specifically, if the kubernetes-object-monitor identifies a matched predicate and sends an uhealthy event, it will not be able to recover and send a healthy event if it misses the corresponding delete event. This is possible if the kubernetes-object-monitor is not running when the delete event occurs. If this occurs, the node will need to be manually unquarantined. Component Health Monitor Steps to Reproduce Modify a K8s object to cause a predicate match that results in an unhealthy event and the k8s-object-monitor-policy-matches node annotation being added. Disable the kubernetes-object-monitor Delete the K8s object that resulted in the predicate match Re-enabl
Prerequisites I searched existing issues Feature Summary Extend the NIC Health Monitor to support GCP TCP Express Offload (TCPXO) devices, enabling link state detection, counter-based degradation monitoring, and syslog-driven failure detection for TCPXO NICs in GPU clusters running on GCP. Problem/Use Case The NIC Health Monitor currently only supports Mellanox/NVIDIA InfiniBand and RoCE devices (mlx5_*). On GCP, GPU clusters use TCP Express Offload (TCPXO) as the primary high-performance networking fabric for distributed training workloads on instances like A3 Mega (H100) and A3 Ultra (H200). TCPXO operates differently from traditional InfiniBand/RoCE — it uses a custom offload engine with a TCPX daemon and GPU-direct datapath, meaning none of the existing mlx5-based sysfs monitoring paths apply. Without TCPXO support, NVSentinel cannot detect grey failures on GCP GPU clusters: No link state monitoring: TCPXO NIC failures, link degradation, and device-level issues go undetected.
Prerequisites I searched existing issues Feature Summary Extend the NIC Health Monitor to support AWS Elastic Fabric Adapter (EFA) devices, enabling link state detection, counter-based degradation monitoring, and syslog-driven failure detection for EFA NICs in GPU clusters running on AWS. Problem/Use Case The NIC Health Monitor currently only supports Mellanox/NVIDIA InfiniBand and RoCE devices (mlx5_*). On AWS, GPU clusters use Elastic Fabric Adapter (EFA) for high-performance inter-node networking, EFA is the primary compute NIC fabric for distributed training workloads on instances like P5 (H100) and P4d (A100). Without EFA support, NVSentinel cannot detect grey failures on AWS GPU clusters, leaving a critical monitoring gap: No link state monitoring: EFA NIC DOWN events, device disappearances, and port state transitions go undetected. No counter-based degradation detection: EFA-specific error counters (packet drops, retransmissions, timeouts) are not tracked, so pre-failure sig
Rate what actually matters after opening a pull request here.
Rate your first-hand PR experience (review speed, maintainer responsiveness, and onboarding ease) to help other contributors.
When evaluating whether to contribute to NVIDIA/NVSentinel, response velocity and maintainer engagement are crucial. GetMerged continuously tracks pull request trajectories, first-comment latency, and code review rounds to help developers avoid submitting pull requests to backlogged repositories.
Currently, maintainers of NVIDIA/NVSentinel acknowledge new external contributions in approximately 2 days+. Out of all submitted pull requests from non-core authors in the last 180-day window, 73.5% were successfully merged into the primary branch.
NVIDIA/NVSentinel has a recorded first-timer success rate of 57.7%. Repositories ranked Welcoming typically provide actionable feedback during code reviews and actively nurture new community contributors.
The initial maintainer response time averages ~2 days+. Keeping PRs scoped to single concerns and ensuring CI checks succeed will optimize review turnaround.
The C-Rank™ index scores repositories on a 0 to 100 scale using an objective formula: external PR merge rates, initial response speed, active maintainer count, and first-time contributor retention. A score of 65.5 places NVIDIA/NVSentinel in the Welcoming tier.
The external pull request merge rate is 73.5%. GetMerged isolates non-core community contributions so external developers get an accurate benchmark of PR acceptance probability.
Yes, NVIDIA/NVSentinel currently has 3 active issue(s) tagged with beginner-friendly labels. You can inspect these directly from the repository issues tab.
All metrics displayed for NVIDIA/NVSentinel are automatically retrieved via the public GitHub API and recalculated daily. Insider pull requests submitted by repository owners or organization members are excluded from merge rate calculations to preserve objective external contributor statistics.