Actions
Architecture #4521
closedWhen generating promises, use a cache for template file
Status:
Released
Priority:
1 (highest)
Assignee:
Category:
Performance and scalability
Target version:
Pull Request:
Effort required:
Name check:
Fix check:
Regression:
Description
For now, we don't cache template file when generating promises, so we read the template each time it is needed to write a promise for a node.
We could cache them in a Map[template id -> template content], trading I/O for memory (and as Linux is quite good in caching, perhaps actually, we don't trade so much I/O to memory).
Actions