Hard Errors when napp-it checks SMART

[DarkNova]

I'm running a napp-it all-in-one with ESXi/OpenIndiana. Using a M1015 card for my drives. I have 5 hard drives in there and I recently added a SSD, a Seagate 600. I noticed that napp-it is logging Hard Errors when SMART is queried for this drive. I know about the problem with the increasing Soft Error counter when querying SMART, but this is the first time I've seen the Hard Errors incrementing.

The command line napp-it is using is: /usr/sbin/smartctl -a -d sat,12 -T permissive /dev/rdsk/....

One weird thing about this is that after the SMART attributes are printed, this command shows:

Read SMART Error Log failed: scsi error aborted command

SMART Self-test Log not supported

Selective Self-tests/Logging not supported

I'm thinking that maybe that "scsi error aborted command" line is making the Hard Error count increase? Each time I run this it seems to generate 4 hard errors.


I played around with it more, and if I change the smartctl command line slightly to: /usr/sbin/smartctl -a -d sat,12 -T normal /dev/rdsk/...

then after the SMART attributes it says:

SMART Error Log not supported

SMART Self-test Log not supported

Selective Self-tests/Logging not supported

then the Hard Errors are not increased.


and then as one final test: /usr/sbin/smartctl -a -d scsi -T normal /dev/rdsk/....

this now doesn't show any SMART attributes, but does show the self-test log properly:

SMART Self-test log
Num  Test              Status                 segment  LifeTime  LBA_first_err [SK ASC ASQ]
     Description                              number   (hours)
# 1  Background long   Completed                   -      29                 - [-   -    -]

Long (extended) Self Test duration: 2280 seconds [38.0 minutes]

So, my questions are, first, is this a problem that these Hard Errors are being generated when SMART is queried? Is it harming anything when this happens? Or should I just ignore them when I query SMART on this drive? Or can I change something with the smartctl configuration to get it to work better with this SSD? Thanks.

 

[gea]

So, my questions are, first, is this a problem that these Hard Errors are being generated when SMART is queried? Is it harming anything when this happens? Or should I just ignore them when I query SMART on this drive? Or can I change something with the smartctl configuration to get it to work better with this SSD? Thanks.

Ignore these iostat messages.
They are driver warnings not real errors (but can indicate future problems).

Also do not rely on Smart too much.
Many disk failures happen without a smart warning, Use higher raid-levels and hotspares to be prepared.

 

[DarkNova]

Thanks. Today I tried the same procedure on a Seagate 600 Pro, connected to a different computer running SmartOS. This also behaves in the same way, with the Hard Error count going up by 4 every time this request is made, so it would appear to have something to do with the firmware design on these Seagate 600s (Pro and non-Pro).