Details
-
Task
-
Resolution: Fixed
-
Major
-
None
Description
The ability to trace the operations on a file will provide a easy way to get a snapshot of changes to the file. This will help with debugging problems which show up after some operation has already occured.
For example if a file corruption is detected when inserting or reading a document, the trace will help determine the operation that touched that part of the file.
We can have a circular buffer which contains the trace of the last N operations and provide an API to dump the trace file. When the upper layer encounters an error, the API can be used to get a trace file.