Conversation
References to 2019 are changed to 2025. The `cis-merge-2019` files are unchanged and needs inspection.
Signed-off-by: Clay Kauzlaric <clay.kauzlaric@broadcom.com>
Signed-off-by: Clay Kauzlaric <clay.kauzlaric@broadcom.com>
Signed-off-by: Clay Kauzlaric <clay.kauzlaric@broadcom.com>
OS version in the manifest is changed to windows2025. The other stemcell tar file fixtures are unchanged because the tests using them don't care about version. Signed-off-by: Clay Kauzlaric <clay.kauzlaric@broadcom.com>
This is required in cases where 'default' network is not used in the GCP project, as is the case for the demo GCP project.
This is required if there is a custom subnet in the network.
Error from packer: `Requested disk size cannot be smaller than the image size (50 GB)`
Since concourse workers and GCP instance created by packer are on the same shared VPC subnet, use internal IP for WinRM access. When external IP is attached, it goes outside the network and that requires firewall rules to enable WinRM access which we don't have permission to create/modify.
AWS uses EC2Launch v2 by default for 2025 AMIs, so this commit changes the way sysprep is invoked.
Signed-off-by: Clay Kauzlaric <clay.kauzlaric@broadcom.com>
These policies are copied from 2019 for now.
Signed-off-by: Brian Upton <brian.upton@broadcom.com>
Windows 2025 base VM uses a 100GB disk instead of 40GB.
This reverts commit 3ede9a6585db472312e119ef72df3abbbc04be68.
Since this director does not have a jumpbox, get rid of the SSH_TUNNEL_* params (ref: cloudfoundry/windows-utilities-release@3bc17e7)
Previously, we were immediately creating a public image for GCP when we built the stemcell. This is not allowed in labs, as we need to publish to a separate account. This commit moves the publishing step to the end of the pipeline, and replaces the previous publishing ruby code w/ a bash script. We have purposefully removed labeling because the dev and public images now live in different accounts. Signed-off-by: Brian Upton <brian.upton@broadcom.com>
* Fix race condition in Windows Update check This commit resolves a race condition in the `install-windows-updates` task where the script prematurely exited the update loop after a reboot because the Windows Update service was temporarily busy or starting up. We now wrap the update check in a try/catch block and retry on failure. Co-authored-by: Cursor <cursoragent@cursor.com> * Improve Windows Update robustness and prevent infinite loop This commit adds several improvements to the `install-windows-updates` task: 1. Prevents infinite loops in CI by introducing a maximum retry counter (10-minute timeout). 2. Strips carriage returns (\r) and trailing whitespace from the captured update count to ensure the strict regex comparison succeeds. 3. Caps the returned update count at 250 in the guest.ps exit code to prevent 8-bit exit code truncation/wrapping. Co-authored-by: Cursor <cursoragent@cursor.com> --------- Co-authored-by: Cursor <cursoragent@cursor.com>
If you don't, you rely on the local cache. Apparently we've been implictly relying on the cache invalidation, but when you trigger builds just often enough to refresh the cache before the KB for patch tuesday is out, you'll NEVER get that patch. Reconstructed timeline: The pattern in the build history explains it. The pipeline ran roughly weekly, and that timing was the key. The Windows Update service automatically invalidates its local cache after ~22 hours and triggers a fresh online scan the next time it's queried. With builds 7 days apart: Build 48 runs (Jun 3) → snapshot captured with fresh WU cache (0 pending) ← 7 days pass, WU cache becomes "stale" → Build 49 runs (Jun 10) → WU service starts from 7-day-old snapshot → detects stale cache → auto-refreshes online → finds 4 updates (those released since Jun 3) For most of the build history (builds 34–48), the ~weekly cadence meant the WU cache was always stale enough to trigger an automatic online refresh. It looked like the pipeline was working correctly, but it was actually depending on cache age to force a rescan. KB5094128 fell through a specific timing gap: 1. Build 48 ran Jun 3 → cache captured, 0 pending 2. June 9 = Patch Tuesday → KB5094128 released 3. Build 49 ran Jun 10 → WU auto-refreshed (7-day-old cache) → but apparently found KB5087068 etc. and NOT KB5094128 yet, possibly because Patch Tuesday updates take 24–48 hours to fully propagate through Microsoft's CDN 4. Build 49 updates the snapshot → cache now shows "just scanned, 0 pending" from the Jun 10 perspective — but KB5094128 still wasn't in it 5. Builds 50/51/52 run the same day or next day → WU cache is only hours old → auto-refresh doesn't trigger → KB5094128 never found So the pipeline wasn't robust — it was relying on cache staleness, and KB5094128 happened to be released during a window where a build ran before the CDN propagation was complete, then the cache got "refreshed" in a way that locked it at a state that excluded KB5094128. Every subsequent same-day or next-day build saw that fresh cache and never went online. -Online skips all this: it always forces a query to Microsoft's servers regardless of cache state, making the behavior deterministic instead of dependent on build frequency and CDN timing. ai-assisted=yes [TNZ-108898]
This should have the same end result as the flag (making sure we don't use a local cache). ai-assisted=yes [TNZ-108898]
|
Note Currently processing new changes in this PR. This may take a few minutes, please wait... ⚙️ Run configurationConfiguration used: Organization UI Review profile: ASSERTIVE Plan: Advanced Run ID: ⛔ Files ignored due to path filters (4)
📒 Files selected for processing (97)
✨ Finishing Touches📝 Generate docstrings
🧪 Generate unit tests (beta)
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 |
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
11 participants
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.