Cloud Garbage Collection Cost Filter for Storage Duration
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Cloud garbage collection processes incur significant costs for customers due to double billing for minimum storage duration, as they are billed for both deleted and new objects, even when objects are still within their storage duration, leading to excessive copy-forward operations and new object creations.
Innovation Solution
Implementing a cost filter phase in cloud garbage collection processes that checks the object creation time against the minimum storage duration, deselecting objects still within this duration from copy-forwarding, thereby reducing unnecessary storage and billing by only copying objects that have exceeded their storage duration.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of energy
If the cloud GC process performs copy-forward operations on objects within minimum storage duration, then storage space is reclaimed, but the customer is double-billed for both deleted and new objects
Solution Approach 1:
The patent applies preliminary action by checking the minimum storage duration of objects before including them in copy-forward operations. The system pre-filters objects based on their retention policy status, ensuring that only objects eligible for deletion (those past their minimum storage duration) are selected for copy-forward. This prevents the harmful effect of double-billing by acting in advance to exclude protected objects from the GC process.
2Volume of stationary object
If the GC process deletes objects still within minimum storage duration, then storage space is freed, but the customer incurs billing for the complete minimum duration
Solution Approach 1:
The system uses feedback by continuously monitoring the minimum storage duration status of objects and adjusting GC operations accordingly. The cloud GC process queries object metadata to determine whether objects have reached their minimum storage duration threshold, and this feedback information directly controls whether copy-forward operations are performed. This feedback mechanism ensures that deletion decisions are made only when it is cost-effective, avoiding billing for objects still within their protected duration.
3Reliability
If multiple GC processes are performed within a minimum storage period, then dead segments are cleaned up, but the number of new objects created increases, leading to substantial additional expenses
Solution Approach 1:
The patent applies preliminary action by implementing a filtering step that checks the minimum storage duration status before GC operations. This pre-check prevents the creation of new objects for segments that are still within their protected period, thereby reducing the number of costly copy-forward operations. The system acts in advance to identify and exclude objects that would trigger additional billing, thus reducing overall storage costs while maintaining data integrity for objects that require protection.
Data Source
AI summary
One example method includes scanning, at a cloud storage site, metadata associated with an object stored at the cloud storage site, fetching, from the metadata, an object creation time for the object, and determining whether the object is out of a minimum storage duration. When the object is out of the minimum storage duration, it is copy-forwarded and then marked for deletion, and when the object is not out of the minimum storage duration, the object is deselected from a list of objects to be copied forward.


