[DocDB] Add memory tracker for temporary memory used by reading WALs

[mdbridge]

Jira Link: DB-18318

Description

Currently we have no tracking of the memory holding WAL records read from disk. (The code that reads these records and allocates the memory is in LogCache.)

This can lead to a TServer crashing due to running out of memory – see #22992: DocDB Xcluster node becomes unreachable due to OOM..

This task works towards fixing this problem by adding a memory tracker for this memory.

Issue Type

kind/bug

Warning: Please confirm that this issue does not contain any sensitive information

• I confirm this issue does not contain any sensitive information.