You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
{{ message }}
This repository has been archived by the owner on Sep 22, 2023. It is now read-only.
Bombastic API and indexer are most heavily affected most likely because of the large objects and index
API consumes memory from loading and searching index, not from GET/POST of objects
Index size in s3 is 1 GB whereas feeding all documents locally gives a size of 33MB. This seems to be caused by missing garbage collection of old segments.
Indexer consumes memory from loading/parsing SBOMs and indexing and also keeping the local index structure in memory
Next steps:
Ensure unused segments are garbage collected
Look at how to use s3 streaming APIs for storing and loading index.
Investigate if index size follows expected size over time if re-feeding the same data many times (no leaks)
Over time it seems that trustification components will have a high memory usage according to OCP. This task is about finding the cause of that.
The text was updated successfully, but these errors were encountered: