Steps to reproduce:

There are no specific steps to perform to trigger this fatal failure. This probably is the worst case scenario as we do not know how to reproduce this issue.

One thing that is worth to mentioning is that we are running 6 LLD worker processes and those processes are heavily loaded.

Result:

All server processes stop processing data. Last message in server log is "malloc_consolidate(): invalid chunk size", nothing more is written to log after that line.

Expected:
Server recovers from this error or at least gives more detailed information about the problem triggering this issue.