Softwares

ClarkDV

New member
What's the best Mac-based software tool for checking the health of a hard drive? I use SMART Utility and Disk Warrior but would like to know how you guys check hard drives.
 

D_R

New member
No Title

@ClarkDV
Of course SMART is the best source for the disk status. But there are more information available than SMART does save. And another problem is to interpret the values of SMART to say: this disk is going to live another 3 Years or fail in a week.

Disk Worrier is not really designed to get a full analysis of your disk. It’s more designed to rescue data. That’s why I suggest to use a proper software for this kind of work. For me, the best software is DriveDx. DriveDx shows you exactly the information about the health of your disk. Here for example the health of mine SSD of my MBPr Mid 2012:
 

Attachments

ClarkDV

New member
I tried out DriveDX earlier today, it looks pretty good. I use DiskWarrior when I suspect a HD is bad but want rock solid confirmation, then I set it to Rebuild and if it starts keeping a count of disk errors while running then I know for certain that that HD should be thrown in the trash.
 

larossmann

Administrator
Staff member
First let's go over some misconceptions here.

diskwarrior is a PARTITION FIXING TOOL not a disk fixer.

It says disk in the name, but that's bullshit.

smart is a piece of shit. use ddrescue. it will go through the entire disk, and if the disk has issues, you already have your clone at the end of the scan.

If you want to recover someone's data on an Apple platform, and make them happy by getting all their apps back and their settings as well as their data,

1) use ddrescue -f -n -a 409600 -b 4096 /dev/baddisk /dev/gooddisk logfile.log to clone the bad disk to a good disk. change 4096 to 512 if the disk's blocksize is 512 and not 4096 - there is information on how to find the blocksize of a disk in linux out there on google.

the reason for step 1 is simple. If we are going to scan the entire disk, and the disk is potentially dying, we might as well recover data as we do that. It is going to skip anything that is a pain in the ass with the command I gave above. anytime that disk is plugged in we should be trying to recover data in a non-destructive way. step 2 will be where we start destroying. :)

2) once that's done use ddrescue -f -d -r 10 -b 4096 /dev/baddisk /dev/gooddisk logfile.log to go over all the parts that were fucked from the first pass. the first pass will ignore what is hard to copy, the second pass will go and fuck the drive to oblivion trying to get those areas once the data from the first pass is safely copied.

3) once your clone is done, take your clone,and bring it over to your hackintosh that has diskwarrior installed... because we're not spending $3000+ on a mac desktop to run diskwarrior. run diskwarrior on the CLONED DRIVE

4) take a third drive -a new one, and install on it the same version of OSX that existed on the failed drive

5) boot into its OS, and use migration assistant to migrate everything over from your diskwarrior'd clone.

the diskwarrior'd clone sometimes will boot, sometimes it will not. even if it does boot, do not use this!!! it has what might be a somewhat fucked up OS and it could act funny down the line. don't be lazy here, no skipping step 4 and 5


diskwarrior should NOT be run on a disk that hasn't had a full scan performed to ensure it doesn't have bad sectors. writing what diskwarrior does to a failing drive is something you will never come back from in some cases, so even if you want todo it the right way later you can't. run diskwarrior on clones of the fucked drive. using diskwarrior on the ORIGINAL BAD DRIVE is like trying to keep a ping pong ball down in a bucket of water using a pencil tip, it's just not going to happen or give good results long term.

If you follow these instructions and you're getting your customers back all of their applications, saved wifi passwords, settings, wallpapers, etc while your competition is stuck with the raw data BS dumped into a folder on the desktop, you will make money hand over fist and have many happy reviews to boot. Particularly from the crowd of people over 40-50. I have had 50somethingyearolds literally break down crying in the store when they had to remember their email server and password.

break down CRYING!!!

you can't make this shit up. When I can migrate data that the other person who worked here couldn't because he was too stubborn to learn this basic process or accept that it works, it makes customers happy that would otherwise be miserable even in cases where we recovered their data. How you recover it, and how quickly they can be back up and running, makes a huge difference.
 

D_R

New member
Hey,

ddrescue and SMART are two different things. Especially when it comes to SSD. Your SSD can be a piece of shit and it might be read with ddrescue without errors, mainly because of the magic of your SSD Controller. But SMART shows you the status of the drive over a long period of time. The idea behind SMART is to have a tool, that can warn you, before a drive gets a fault so you can rescue your data.
 

ytesfay80

New member
larossmann
on step 2 you say " once that's done use ddrescue -f -d -r 10 -b 4096 /dev/baddisk /dev/gooddisk logfile.log to go over all the parts that were fucked from the first pass."

Do you use that command on the cloned drive or on the original bad drive?
 

larossmann

Administrator
Staff member
The original bad drive. You do two passes - one to get the easy to get data, the second to get the pain in the ass to get data. The idea is to not stress the drive on the first pass using up what little life it has left trying to grab a few sectors when you could be using that life to grab the other 99% of the data.
 
Top