Yes, it is definitely the oom-killer that has been terminating the database kernel. I would now suggest the following:
1.
Find out, if the memory usage of the database process 'kernel' is increasing over time or if there is an upper limit after some days of database usage.
One method to determine the memory usage is to start the MaxDB Analyzer and monitor the heap usage. The database kernel memory should be mostly allocated by the defined cache_size parameter plus some more memory on top that gets allocated when needed.
2.
In case the memory usage seems to grow more and more over time you should consider upgrading your MaxDB 7.8.01.14 to a recent version of 7.8 as you might be experiencing a memory leak in the database kernel. If it is really a memory leak, then increasing the memory for the VM would only temporarily help.
3. Maybe the MaxDB cache_size is configured too big or the VM has not enough memory available, so adjusting one of these parameters would help (unless it is a memory leak, of course).
Thorsten