Server fall

[TODDLT]

Tonight I went to add a new hard drive into my case to pre-clear, and update to the latest rc5.

 

I shut down and sat the server on the floor and was performing the update when my 7 YO knocked the server over.  It fell on a padded/carpeted floor with a pretty loud thud.  It wasn't elevated, only tilted over, but it's a tall heavy box so still hit hard.

 

I had my mild stroke, put the new drive in and rebooted.  Everything seems to be fine, drives all recognized. 

 

However, is there any utility I should run that would adequately check the drive for damage?  Would a short or long smart test of each be beneficial?

 

Old hard drives used to be a lot more susceptible to shock, but still nervous.

[lionelhutz]

They're probably OK since they were powered off.

 

Maybe monitor the SMART reports to ensure nothing is going bad.

[PeterB]

They're probably OK since they were powered off.

 

Maybe monitor the SMART reports to ensure nothing is going bad.

 

Indeed - check the drive specifications for shock tolerance when powered off - it can be quite high.  If the drives had been spinning, this episode might have had a completely different conclusion.

[TODDLT]

Cool, thanks to both of you for a little piece of mind 

[marcusone]

Run a parity check asap. The drives can still be damaged and only way to know for sure is to work them. Also listen for any abnormal clicking or other noises not there before the incident.

 

Sent from my SGH-I727R using Tapatalk 2

[c3]

Test them, but move with purpose. I put this here more for the next time, since i expect you have already brought it online.

 

Run a smart short test for each drive

Collect full report for each drive

Run smart long test for each drive

Online and parity check without correction (you don't want change parity if non parity has trouble)

compare smart report for growing errors

[TODDLT]

I am running a parity check now, so far so good at about 24% complete.  If it comes back clear, I will close this.

 

I also did a short smart test on all drives in the array.  Everything looked pretty normal. 

 

My two Samsung drives have an odd ball number, but they are similar for both drives and found similar questions asked by others online.  So I'm guessing this is just something specific to Samsung and not worth worrying about.

181 Program_Fail_Cnt_Total  0x0022  100  100  000    Old_age  Always      -      20335034

181 Program_Fail_Cnt_Total  0x0022  099  099  000    Old_age  Always      -      22178014

 

Similarly on both Seagate drives (one of which wasn't in the box when it tipped), SER, RRER, and HER raw numbers are very high, but doing some online research this again appears to be specific to how Seagate uses these numbers.

 

1 Raw_Read_Error_Rate    0x000f  117  099  006    Pre-fail  Always      -      156637184

7 Seek_Error_Rate        0x000f  100  253  030    Pre-fail  Always      -      764549

195 Hardware_ECC_Recovered  0x001a  036  013  000    Old_age  Always      -      156637184

 

(this is the new drive, and doesn't have HER)

1 Raw_Read_Error_Rate    0x000f  116  100  006    Pre-fail  Always      -      108795448

7 Seek_Error_Rate        0x000f  061  060  030    Pre-fail  Always      -      1489340

 

I am running a parity check now, so far so good at about 26% complete.  If it comes back clear, I will close this.