Preventing HD Errors/Corruption. Possible?

[danshaw]

In our company I see every now and again a laptop that has "unexplained" issues. Most of the time when I run a Verify Disk with Disk Utility I find that there are disk issues ranging from corruption to directory and file inconsistencies. I've read that this can happen over time.

To solve this I boot into the recovery partition, run first aid on the disk, repair it, and send the user on their way. All is well in their world again.

I'm wondering if there is any way to prevent these issues from happening over time. Does anyone run anything on their users computers that essentially keeps the computers clean and error free so that issues don't build up over time? Not sure if running diskutil from the command line occasionally would help or if it will only cause more issues.

[dferrara]

Yes we see this a lot. HFS/+ is very, very old. I think we are all basically waiting for Apple's new filesystem to become mature enough for daily use. I have heard a lot of good things about OpenZFS, which is probably more viable for servers and RAIDs than client systems.

[analog_kid]

Same here: out of 1800 clients we regularly have 125-150 with disk verification issues. I'll second that HFS+ needs to be taken back behind the barn... luckily the end is in sight with Apple's new filesystem that's in beta currently.

Based on this post I created both a Self Service job (for end-users) and a fully automated solution (for lab computers) for repairing the disk in single user mode. Ignore my comment in that post as it seems in later versions of Yosemite, everything works peachy. Tested on OS 10.8-10.11.

Modifications I made were creating an Extension Attribute for tracking the number of times a repair is attempted so we can report on machines that may have deeper disk issues. The repair count EA is also used as a kill switch in our automated version of this so that a machine doesn't endlessly try to repair a disk that's not repairable via fsck.

Good luck!