Conversation
Signed-off-by: Wenqi Li <wenqil@nvidia.com>
Dependency Review✅ No vulnerabilities or license issues or OpenSSF Scorecard issues found.Scanned FilesNone |
WalkthroughTwo new JSON manifest files added for Holoscan release 4.0.0. The artifacts.json contains release metadata including debian and wheel versions, base images, build images across multiple platforms (dgpu, igpu, cpu), and health probe URLs. The artifacts-cu12.json provides the same structure for CUDA 12 configurations, covering release 4.0.0 and historical versions down to 3.3.0. Changes
Estimated code review effort🎯 2 (Simple) | ⏱️ ~12 minutes Possibly related PRs
Suggested reviewers
🚥 Pre-merge checks | ✅ 2✅ Passed checks (2 passed)
✏️ Tip: You can configure your own custom pre-merge checks in the settings. Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out. Comment |
![coderabbitai[bot]](https://avatars.githubusercontent.com/in/347564?s=60&v=4){kind=small}
There was a problem hiding this comment.
Actionable comments posted: 2
🧹 Nitpick comments (1)
releases/4.0.0/artifacts-cu12.json (1)
26-29:grpc-health-probev0.4.38 is pinned across all entries — a newer version is availableThe latest grpc-health-probe release is
v0.4.39(and further releases exist up to v0.4.45), while all entries pinv0.4.38. Pinning is generally safe and v0.4.38 is a valid, confirmed release, but consider evaluating whether a newer version should be adopted for the 4.0.0 manifest.🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed. In `@releases/4.0.0/artifacts-cu12.json` around lines 26 - 29, The "health-probes" entries currently pin grpc-health-probe to v0.4.38; update the two URLs under the "health-probes" key to point to the newer release (e.g., change the version segment from v0.4.38 to v0.4.39 in the linux/amd64 and linux/arm64 URLs), then run any release validation you use (download/sha/checks) to confirm the new binaries; alternatively, if you intend to adopt a different newer patch (v0.4.40+), replace v0.4.38 with that chosen tag consistently in both URLs.
🤖 Prompt for all review comments with AI agents
Verify each finding against the current code and only fix it if needed.
Inline comments:
In `@releases/4.0.0/artifacts-cu12.json`:
- Around line 10-24: The build-images entries ("build-images" -> "igpu", "dgpu",
"cpu") currently reference the staging registry prefix
"nvcr.io/nvstaging/holoscan/holoscan:v4.0.0.0-cuda12-*" and must be updated to
the production registry paths; locate the image strings under the keys igpu,
dgpu, and cpu and replace the "nvcr.io/nvstaging/..." prefixes with the correct
production registry/image names for v4.0.0.0-cuda12 (ensuring each image tag
stays the same, e.g., replace
nvcr.io/nvstaging/holoscan/holoscan:v4.0.0.0-cuda12-igpu with the corresponding
production registry path).
In `@releases/4.0.0/artifacts.json`:
- Around line 9-18: The manifest's build-images entries use the staging registry
host "nvcr.io/nvstaging" (see keys "build-images" -> "dgpu"/"cpu" image
strings), which will be unreachable to most users; replace those staging image
references with the production registry namespace (e.g., change
"nvcr.io/nvstaging/holoscan/holoscan:v4.0.0.0-cuda13" to the agreed production
host such as "nvcr.io/nvidia/holoscan/holoscan:v4.0.0.0-cuda13" or
"nvcr.io/nvidia/clara-holoscan/holoscan:v4.0.0.0-cuda13") and ensure the
corresponding images are promoted to that production registry before this
manifest is merged.
---
Nitpick comments:
In `@releases/4.0.0/artifacts-cu12.json`:
- Around line 26-29: The "health-probes" entries currently pin grpc-health-probe
to v0.4.38; update the two URLs under the "health-probes" key to point to the
newer release (e.g., change the version segment from v0.4.38 to v0.4.39 in the
linux/amd64 and linux/arm64 URLs), then run any release validation you use
(download/sha/checks) to confirm the new binaries; alternatively, if you intend
to adopt a different newer patch (v0.4.40+), replace v0.4.38 with that chosen
tag consistently in both URLs.
| "build-images": { | ||
| "igpu": { | ||
| "jetson-agx-orin-devkit": "nvcr.io/nvstaging/holoscan/holoscan:v4.0.0.0-cuda12-igpu", | ||
| "igx-orin-devkit": "nvcr.io/nvstaging/holoscan/holoscan:v4.0.0.0-cuda12-igpu" | ||
| }, | ||
| "dgpu": { | ||
| "x64-workstation": "nvcr.io/nvstaging/holoscan/holoscan:v4.0.0.0-cuda12-dgpu", | ||
| "igx-orin-devkit": "nvcr.io/nvstaging/holoscan/holoscan:v4.0.0.0-cuda12-dgpu", | ||
| "sbsa": "nvcr.io/nvstaging/holoscan/holoscan:v4.0.0.0-cuda12-dgpu", | ||
| "clara-agx-devkit": "nvcr.io/nvstaging/holoscan/holoscan:v4.0.0.0-cuda12-dgpu" | ||
| }, | ||
| "cpu": { | ||
| "x64-workstation": "nvcr.io/nvstaging/holoscan/holoscan:v4.0.0.0-cuda12-dgpu" | ||
| } | ||
| } |
There was a problem hiding this comment.
Same nvcr.io/nvstaging/ concern as artifacts.json — all 4.0.0 build images are staging
igpu, dgpu, and cpu build images for 4.0.0 all point to nvcr.io/nvstaging/holoscan/holoscan:v4.0.0.0-cuda12-*. Same pre-production concern as noted in artifacts.json: these need to be replaced with production registry paths before this manifest is served to end users.
🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed.
In `@releases/4.0.0/artifacts-cu12.json` around lines 10 - 24, The build-images
entries ("build-images" -> "igpu", "dgpu", "cpu") currently reference the
staging registry prefix "nvcr.io/nvstaging/holoscan/holoscan:v4.0.0.0-cuda12-*"
and must be updated to the production registry paths; locate the image strings
under the keys igpu, dgpu, and cpu and replace the "nvcr.io/nvstaging/..."
prefixes with the correct production registry/image names for v4.0.0.0-cuda12
(ensuring each image tag stays the same, e.g., replace
nvcr.io/nvstaging/holoscan/holoscan:v4.0.0.0-cuda12-igpu with the corresponding
production registry path).
| "build-images": { | ||
| "dgpu": { | ||
| "x64-workstation": "nvcr.io/nvstaging/holoscan/holoscan:v4.0.0.0-cuda13", | ||
| "sbsa": "nvcr.io/nvstaging/holoscan/holoscan:v4.0.0.0-cuda13" | ||
| }, | ||
| "igpu": {}, | ||
| "cpu": { | ||
| "x64-workstation": "nvcr.io/nvstaging/holoscan/holoscan:v4.0.0.0-cuda13" | ||
| } | ||
| } |
There was a problem hiding this comment.
Staging registry in 4.0.0 build images — confirm images will be promoted before this manifest is consumed
All 3.x entries reference the production registry (nvcr.io/nvidia/clara-holoscan/holoscan), but the 4.0.0 build-images reference nvcr.io/nvstaging/holoscan/holoscan:v4.0.0.0-cuda13. Users without access to the staging registry will fail to pull these images when using this manifest. Please confirm the staging images will be promoted to the production namespace (nvcr.io/nvidia/holoscan/holoscan or nvcr.io/nvidia/clara-holoscan/holoscan) and that this file will be updated with the production tags prior to merge.
🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed.
In `@releases/4.0.0/artifacts.json` around lines 9 - 18, The manifest's
build-images entries use the staging registry host "nvcr.io/nvstaging" (see keys
"build-images" -> "dgpu"/"cpu" image strings), which will be unreachable to most
users; replace those staging image references with the production registry
namespace (e.g., change "nvcr.io/nvstaging/holoscan/holoscan:v4.0.0.0-cuda13" to
the agreed production host such as
"nvcr.io/nvidia/holoscan/holoscan:v4.0.0.0-cuda13" or
"nvcr.io/nvidia/clara-holoscan/holoscan:v4.0.0.0-cuda13") and ensure the
corresponding images are promoted to that production registry before this
manifest is merged.
1 participant
Summary by CodeRabbit