Hard Drive Crash: Best Practices

It happens, the inevitable crash. Happened to me just last week. I luckily do perform backups, and have most of my desktop replicated to my laptop, however, even bouncing lost items back leaves you, in many cases with a new OS and re-installing and configuring the workstation. My question to the forum:

“What are some proven methodologies you have used that allow users to get up and running swiftly, upon the unfortunate incident of your hard drive taking a hard left into a median, when it should have simply stopped texting while driving and NOT crash.” [Excuse my little analogy there :slight_smile: ]

Look forward to any solutions, hopefully we can fight this evil :slight_smile:

Recovery is a pain in the butt and is almost unavoidably long. The biggest thing we do it to try to keep a list of software that needs to be installed, and how it should be configured. The hardest part is not clicking the OK button a thousand times, but remembering what you need to run to get to that OK dialog. Actually, we wrote a C# app a long time ago that automates the installation of 90% of the software used by our teams, and it has different configurations depending on whether you are an artist, engineer, producer, etc.

Personally, I’ve never* had a hard drive fail that didn’t give me some indication that it was on its last legs days in advance. Don’t ignore warning signs, and get a new HD before it fails.

  • I had a power supply go bad that took out my motherboard, gfx card, HD, etc. and I had to replace the whole thing. Since the platters were fine on the disk, I just got on EBay, found the same exact drive, swapped the controller cards and kept using it. I doubt SSD drives are going to be that friendly…

After posting this I started reflecting on it more, and I was thinking that may be the best approach right now. Collecting a list/folder of my preferred setup along with the required installation executables and a detailed document explaining the step-by-step setup required for other third party or related apps and associated path variables. Some could be automated, and for a studio that would make sense. I might look into scripting some of it, just as an exercise, since in my case its for personal use.

On my actual HD failure, pretty sure the HD failed due to the move. Three days in a moving truck likely being jostled around makes for a not-so-happy drive.

Some of the most annoying hassles involve configuring the little stuff such as “which program is the default for opening .jpg files?”

This won’t get you 100% back up and running, but I use it a lot to get my basics back in order with little hassle:
http://ninite.com/

But yes, it’s the little things like file associations that really rankle until you get them set up again.

Well the best solution is on the hardware side of things imo: mirroring (multiple harddrives) - or am I missing something?
Even if a drive is physically damaged, you can restore everything from the mirror HDD(s).
Also: it’s a good idea to mirror all your data to a server not located in your offices (in case of fire or natural disasters)

I actually did have that done, well at least to a server within my apartment (I don’t have access to something external), and mostly backed up, not a full mirror. I was more questioning a proven process to get the system and the programs associated with it up and running once the new OS was installed.

Example: Hard drive crashes. You fortunately have all of your data backed up to the server etc. You still need to replace your main HD and install a new OS. Upon which you will need to re-install your associated programs as well as configure their settings, and as I’m learning now, once again configure any third party associations for your pipeline within those programs. Many of which are command line driven installations.

That was sort of my thinking. When it kicked, I had a lot of personal settings and third party applications installed that I really had not documented (Years of picking and trying different things), so I’m reaching for things and finding they’re not there. Basically, what I’m learning, mirror and document a step-by-step process along with a list of the programs used and keeping an INSTALLATIONS or something similar folder with the required packages you use someplace safe is a good start. Also, secure your computer when moving, and don’t toss the ball too close to your workstation. The “Mutt” cares more about destroying the ball then bumping my computer :slight_smile:

dd if=/dev/sda of=/dev/sdb :):

Or the use some imaging software could help? Clonezilla, Acronis true image etc. I’ve thought about using a virtualization layer for easy management of snapshots and to be able to restore to any hardware. Works fine as long as you’r not doing anything 3D heavy…
But it’s really nice to have a image of your whole system that you can transfer with you and boot on any machine basically.

EDIT: If you are using linux for your workstation you can also do “RAID1 over network” with DRBD.