It might be that you want output to an IFS file be in sync with journaled output to DB2 tables, and only at a commit-point that the data actually gets written to the IFS file? I did it this way...
Some basic steps I performed in order to have git running locally in an isolated network area for a private project. It seemed to be as straigtforward as dreamed.
This is a solution I wrote when an existing tool was suspected of causing a memory leak in a Windows svchost process. This script scans a list of folders, and sends the found PDFs to a printer with the same name as the folder, unless mentioned otherwise in the .ini-file (see below).
Sync attempts result in errors like
The reason for this is that you need some more directives for Apache.
ZFS has, due to it's time stamps in metadata blocks and the technique of copy-on-write the advantage of heaving a great snapshot implementation. There are several scripts available which provide you with backup options using snapshots (full, incremental, differential).
In this example asp01 is the main data pool, where asp01b will be a pool created on a backup disk (erasing anything else which was already on that disk).