* Inital changes needed for backoff 0.4 to work
* Update backoff versions, fix BackoffError:Transient fields, other uses
* Format
* Removed redundant field name
* Improved backoff update changes
* Update backoff update
* Revert
* Changed to using Error::transient function
* Abstract node disposal strategy
* Cleanup + lint
* Handle possibile scalesets being in resize state
* Setting the size is still exposed via CLI, we don't want to break that functionality
* PR comments
* Release 5.1.0
* Update CHANGELOG.md
Co-authored-by: Joe Ranweiler <joe@lemma.co>
* Update CHANGELOG.md
Co-authored-by: Joe Ranweiler <joe@lemma.co>
* Update CHANGELOG.md
Co-authored-by: Joe Ranweiler <joe@lemma.co>
* Prevent deletion of the repro VM on failure for debugging.
Co-authored-by: Joe Ranweiler <joe@lemma.co>
Add a setup script-specific timeout of 59 minutes. This is just shorter than the service-side `NODE_EXPIRATION_TIME` which otherwise garbage collects nodes whose setup scripts are stuck or taking too long.
With this change, the high-level cause of the timeout is clear, instead of the closest error being something indirect, like "node reimaged during task execution".
- Add `onefuzz::memory::available_bytes()` to enable checking system-wide memory usage
- In managed task worker runs, heuristically check for imminent OOM conditions and try to exit early
* Initial progress to adding a auto scale resource
* auto scale API is ready
* When creating a scaleset, add an autoscale resource to it as well
* Auto scale is correctly linked with scaleset
* 🧹
* Lint
* Cleaned up
Refactoring check-pr.py to extract the logic of downloading the binaries
refactoring integration-tets.py to split the logic of setup, launch, check_result and cleanup
* draft attempt at adding scaling protection
* Service can now control scaling protection policy on VM instances
* Improve logging a bit
* draft attempt at adding scaling protection
* Service can now control scaling protection policy on VM instances
* Improve logging a bit
* Error message was missing info
* Linter
* Don't schedule work if we can't protect the node
* Last of the linter changes
* Update yanked block-buffer from 0.10.0 to 0.10.1
* update yanked crossbeam-utils from 0.8.5 to 0.8.7
* update yanked crossbeam-utils from 0.8.5 to 0.8.7
Co-authored-by: stas <statis@microsoft.com>
* Fixing VMSS Re-Image 7-Day Timer
* Updating use of TimeStamp to created_at
* Renaming.
* Updating field name.
* Removing test chagne.
* Updating query to work if init_at entry does not exist yet.
* Changing timer for testing.
* Adding field comment.
* Formatting models.py
* Fixing where save is called.
* Adidng logging.
* Removing logging. Ready for merge.
* Update src/pytypes/onefuzztypes/models.py
Co-authored-by: Joe Ranweiler <joe@lemma.co>
* Formatting.
* Updating datetime.
* Testing after datetime change.
* Removing test.
Co-authored-by: nharper285 <nharper285@gmail.com>
Co-authored-by: Joe Ranweiler <joe@lemma.co>
