SSD's were acting up, seem to work fine after Clean?

Blazestorm

Supreme [H]ardness
Joined
Jan 17, 2007
Messages
6,940
I had these two GSkill Falcon 128GB in RAID-0 for a long time. Just yesterday they were acting weird, computer was freezing, Intel was saying the array was degraded. Just weird stuff. I followed the awkward GSkill procedure for wiping them (Install jumper pins, downgrade to old firmware, run clean command, remove jumper pins, boot to DOS, upgrade to new firmware). And seems to have brought my performance back.

Anyone have any ideas for testing these more thoroughly?

http://imgur.com/a/4MQt3
 
Since the drives are working fine now, it's probably not a problem with the drives themselves. However, if the raid performance was diminished to a substantial degree, it could some problems. I would submit that you should probably add in some more overprovisioning and just SE them a couple times a year.

You should probably change SSDlife to display SMART data in decimal and not hex. If you look at the second drive, you'll see that there were some read failures on one of the drives. It looks like you did the destructive flash, so much of the SMART data was reset.
 
Yea, I was having issues even booting into Windows before. And it seems like that 2nd drive is the one acting up. Since they were in RAID it was hard to tell, I restored my WHS backup to a 1TB drive before I started messing with these.

Didn't even consider checking the SMART data before I wiped them.

Any other tests I can run to stress them / see if that 2nd drive is going to cause more problems in the future?
 
Maybe -- but the problem would be the fact that by stressing the drive to see if more errors occur, you'd be causing more wear on the drive and possibly create more issues in the process.

So... I recommend setting the array back up, leaving some over provision space, and seeing how it goes. Since the SMART data got wiped, I can't tell how much was being written before, but I'd guess that the array was pretty slow. Without TRIM, write amplification was pretty high, magnifying drive wear to a degree. If you put the array back and leave some unpartitioned space on the set, it will help control performance degradation and write amplification. I'll bet that unless one of the drive's NAND devices is going bad, everything will be fine for a while.

Keep decent backups.
 
Maybe -- but the problem would be the fact that by stressing the drive to see if more errors occur, you'd be causing more wear on the drive and possibly create more issues in the process.

So... I recommend setting the array back up, leaving some over provision space, and seeing how it goes. Since the SMART data got wiped, I can't tell how much was being written before, but I'd guess that the array was pretty slow. Without TRIM, write amplification was pretty high, magnifying drive wear to a degree. If you put the array back and leave some unpartitioned space on the set, it will help control performance degradation and write amplification. I'll bet that unless one of the drive's NAND devices is going bad, everything will be fine for a while.

Keep decent backups.

I did check the SMART data but not very closely. They had ~6 months of usage ( I had just wiped them back then too ) and said 65% life vs. 98/100%. Had about 300 power-on's, but besides that everything appeared ok.

The array was slowish, but mainly just the whole Windows locking up and crashing / being unusable was the issue. Slowing down is still pretty fast for SSD's.
 
Well, I'm curious to know whether everything works fine after the wipe.
 
Back
Top