problems after vacuum full

Hi,

We have a virtual machine (VMware ESXi 6.5 U2):
OS: 4.4.0-96-generic #119-Ubuntu SMP Tue Sep 12 14:59:54 UTC 2017 x86_64
x86_64 x86_64 GNU/Linux
PG: PostgreSQL 9.6.5 on x86_64-pc-linux-gnu, compiled by gcc (Ubuntu
5.4.0-6ubuntu1~16.04.4) 5.4.0 20160609, 64-bit

A database was working for a few months and I did a vacuum full on one
table. A table had about ~160 GB of data and the same amount of indexes.
After vacuum full was successful a table was emptied - SELECT on it
returned 0 rows, but not an error.

During the OS analysis, we discovered differences between du and df
commands that indicated there were deleted files with open handlers.
Command 'lsof' displayed postmaster as PID holding the handlers.

We would like to know if somebody had similar issues? We are afraid that
it would happen to us again. We are aware of the fact that we should
make an upgrade, but we are currently unable to do so.

We would appreciate if you could give us any suggestions.
Best regards

Aleksandra Angielska