I did an experiment yesterday on trying to improve the performance of my backups. I think I can categorize it as successful but it's only a marginal improvement. What I've done is setup an old workstation of mine as a backup proxy. The job is backing up all but my most unimportant VMs and you can see that it started out strong but gradually lost its steam.
Since this is an experiment and because this machine is old and uses server RAM (ECC), I am currently very limited in that department at just 4GB. I noticed that over the course of the job (which was 8h30m long) the memory utilization went up, albeit very slowly.
I wasn't able to capture a graph of the memory usage because, as I found out this morning, I hadn't setup monitoring for this machine in PRTG.

If I invest a little bit of money in getting this machine more memory (I'm thinking in the 16GB range) should I see even stronger and more sustained performance over the course of the job? Or are these two metrics merely coincidental?
