Repository navigation
fcvm-fleet: host and tile AMIs from ejc3/fcvm main 728be8c3 (clouds and weather) - #713
Conversation
…nd weather) Host fcvm-fleet-host-x86_64-728be8c3-20261027 (ami-07ea807a8a9b4c3fa) and tile fcvm-fleet-tile-x86_64-728be8c3-20261027 (ami-098e582592b24a025), replacing the ce08eee9 images from #711. 728be8c3 contains ce08eee9 and adds the renderer's fluffier clouds and weather keys (ejc3/fcvm#1383).
|
No actionable comments were generated in the recent review.
 |
| Layer / File(s) | Summary |
|---|---|
Update host and tile AMI defaults fcvm-fleet.tf |
The host AMI default changes to ami-07ea807a8a9b4c3fa, and the tile AMI default changes to ami-098e582592b24a025. Both comments identify revision 728be8c3. |
Priority:
Estimated code review effort: 1 (Trivial) | ~3 minutes
Change: Feature
Possibly related issues
- ejc3/aws#712: This issue tracks staging AMI updates to the same host and tile defaults.
- ejc3/aws#700: This issue concerns an earlier update to the same host AMI default.
Possibly related PRs
- ejc3/aws#698: This PR updates the same host and tile AMI defaults.
- ejc3/aws#699: This PR updates the host AMI default in the same image progression.
- ejc3/aws#659: This PR updates earlier fleet host and tile images.
Merge Risk: 
eae19
-
No actionable merge-blocking risk is identified in this AMI defaults update.
Pre-merge checks: 
-
Check name Status Explanation Description Check Check skipped - CodeRabbit’s high-level summary is enabled. Title check The title clearly identifies the host and tile AMI updates and the source fcvm commit. It accurately reflects the primary change. Docstring Coverage No eligible functions found in the changed files. Skipping docstring coverage check. Docstring coverage is scoped to functions touched by this diff with cyclomatic complexity of at least 2. Documentat… Linked Issues check Check skipped because no linked issues were found for this pull request. Out of Scope Changes check Check skipped because no linked issues were found for this pull request. Full details: Docstring Coverage
-
Explanation
No eligible functions found in the changed files. Skipping docstring coverage check. Docstring coverage is scoped to functions touched by this diff with cyclomatic complexity of at least 2. Documentation-only edits and deletions within surviving functions are included; unchanged diff context and fully deleted functions are excluded. Analyzed 0 functions across 0 files. (1 skipped: 1 unsupported.)
-
Finishing Touches
-
🧪 Generate unit tests (beta)
-
- Commit to this branch
- Create a new PR
-
- Autofix · Keep fixing CodeRabbit findings and required CI, and resolving merge conflicts
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 help to get the list of available commands.
Codex Review SummaryThis comment shows the latest Codex review activity on this pull request.
ℹ️ About Codex in GitHubYour team has set up Codex to review pull requests in this repo. Reviews are triggered when you
Codex reacts with 👀 while any review is running, comments if it has suggestions, and reacts with 👍 once all reviews finish with no findings. |
|
@codex security review |
|
@coderabbitai review |
|
@greptileai review |
✅ Action performed
|
🛡️ Codex Security ReviewSecurity review completed. No security issues were found in this pull request. Reviewed commit: Only the user who started this review can view the report in Codex. ℹ️ About Codex security reviews in GitHubThis is an experimental Codex feature. Security reviews are triggered when:
Once complete, Codex will leave suggestions, or a comment if no findings are found. |
|
Ready at eae19b6: independent HIGH code/security/artifact review, required AWS CI, exact-head Codex code/security reviews, and actual CodeRabbit review pass; no unresolved threads. Greptile did not respond and is not counted as approval. Verified both private fleet-owned AMIs, builder/source lineage, all 26 recipe bundle files against source 728be8c3, completed store parts before imaging, existing 87/87 tile GPU checks and host warm-launch logs. Exact upstream main CI was cancelled before jobs, so it is not represented as a green full source run. The existing build/runtime evidence was reviewed directly. The host follow-up #716 is stacked here; it retains this tile image and advances only the host's boot-prefetch recipe. These image pins apply on subsequent launches. |
|
Applied at merged main 156cf71. The reviewed full plan applied successfully (2 added, 6 changed, 0 destroyed), and the full 2,526-resource follow-up returned exit 0 / no changes. Live latest launch templates now use the retained 728be8c3 tile image ami-098e582592b24a025 (spot/on-demand version 7; pool version 4) and #716's final 7f5fb7e6 host image ami-0f1f1af48e2312360 (version 10). The coordinator remains healthy on its existing task definition revision 47. Existing hosts were not forcibly recycled; the pins apply on subsequent launches. |
Host and tile AMIs from ejc3/fcvm main 728be8c3, whose renderer draws fluffier clouds and the scene's weather keys (ejc3/fcvm#1383). The coordinator's director already publishes weather per stop (ejc3/fcvm#1389, deployed to staging by the fcvm CI); renderers on the current images ignore those keys, so staging shows neither change until these images run.
fcvm_fleet_host_amis)fcvm_fleet_tile_amis)728be8c3 contains ce08eee9, so these images carry everything #711's did (ejc3/fcvm#1361 the H.264 bitstream restriction, #1380 every tile inside its frame slot, #1382 the restored guest's clock re-read, #1370 the baked agent snapshot and boot prefetch) and supersede #711, #708 and #710. New since ce08eee9 in the renderer: ejc3/fcvm#1383 (clouds: a soft density fringe, four deeper scattering orders, powder from the sun's optical depth, sky light through the cloud above; weather: cover, towering cumulus, crepuscular rays, snow line). On the staging wall's layout (16 tiles of 640 x 360 at 420,448 pixel-samples) the slowest tile measured 13.6 to 19.5 ms with default weather and up to 19.9 ms with towers, against the controller's 23.3 ms target.
Built with ejc3/fcvm
fleet/ami/build-ami.sh:build(fcvm at 728be8c3,make setup-gpu, renderer image, warm VMs: root warm-up first frame 13.1 s, 6163 files, 335 MiB to prefetch) andbake, store parts understore/fcvm-fleet-host-x86_64-728be8c3-20261027/(16 PUTs, all 200).units,image(the renderer export from the host build) andbake.fleet-renderer-testin that image on its T4: 87 checks passed, its weather check's hash equal to the one two fcvm VMs on another box printed.measure-boot.shon the new images (spot): tile first frame over HTTPS 44.7 s after the accepted request (26 s of capacity refusals not counted); host GPUs on vfio-pci 78.3 s and store restored from its own prefix 88.3 s after the request (16 parts in 10.0 s), one GPU VM to first frame 5.9 s from the baked snapshot (snapshot: hit). Both instances terminated; both builders terminated.Tested:
python3 -S -B scripts/test-fcvm-fleet.pyOK (51 tests),terraform fmt -check.Plan: only the two AMI defaults change; running hosts and stopped tiles keep their images until the fleet next sleeps and wakes.
Summary by CodeRabbit