What makes less hardware "pressure": copy 100 items or zip them?
Posted: 23 Feb 2025 10:53
Question is theoretical, but since i don't dwell on "ask whatever"-platforms, like Reddit-sorts, i simply don't know where else to ask this. Yet question appeared on my scope while making very often backups via XY scripts, which have zip support = so main goal is quite in scope of XY field.
So does anyone have knowledge what will make less I/O and hardware pressure: a copy of say 100 items(which means +100 write ops to create new items) OR zipping them in single file?
Example. A backup that happens before each run of particular app and makes safety copies of important items. Sometimes it may be only couple times per day, sometimes more than a dozen. So counting those number of operations, plus each day, yet all that "just to be safe" - led me to thought = maybe making a zip of those items is a better choice? Cause at least by looks, it's way less excessive I/O to create one item for whole list, instead of whole list mirror on each backup.
So does anyone have knowledge what will make less I/O and hardware pressure: a copy of say 100 items(which means +100 write ops to create new items) OR zipping them in single file?
Example. A backup that happens before each run of particular app and makes safety copies of important items. Sometimes it may be only couple times per day, sometimes more than a dozen. So counting those number of operations, plus each day, yet all that "just to be safe" - led me to thought = maybe making a zip of those items is a better choice? Cause at least by looks, it's way less excessive I/O to create one item for whole list, instead of whole list mirror on each backup.