Hard Drive Burn-In Testing - Discussion Thread

[Mirfster]

I personally waited until after the long tests were done before starting the badblocks.

 

[bmcclure937]

I am planning to burn in my new drives and run these tests on Friday evening. Running a 50 hour memtest86+ before I start looking at the HDDs. I have 2x6TB WD Red drives. Based on what others are saying, I expect this to take over 50 hours to complete.

 

[bmcclure937]

I am over 50 hours into my badblocks test for 2x6TB HDD.

Code:

badblocks -b 4096 -ws /dev/adaX

Has anyone else run badblocks on 6TB drives? How long can I expect these tests to take?

Thanks!

 

[nightshade00013]

I am over 50 hours into my badblocks test for 2x6TB HDD.

Code:

badblocks -b 4096 -ws /dev/adaX

Has anyone else run badblocks on 6TB drives? How long can I expect these tests to take?

Thanks!

It will depend on the RPM of the drive. My father had some 4tbWD Red's and it took around 85 hours for a complete run through of all 4 passes. I scored some 4TB HGST NAS drives for twenty bucks less than the WD Red's and it took around 78 hours.

To get an estimate of a pass time do a

Code:

smartctl -a /dev/adaX

Look for the section about smart long test length. Take that times two, once for read and once for write and then times four since badblocks will do four cycles and you will get an approximate timeframe in minutes for the entire test. It's not perfect but will give you a timeframe to start watching for it to be done.


So for instance on my HGST drives I get

Short self-test routine
recommended polling time: ( 1) minutes.
Extended self-test routine
recommended polling time: ( 560) minutes.

560 x 2 = 1120 x 4 = 4480 / 60 = 74 hours. It took just a few hours longer than that since write is a little slower to read but was within 4 hours of the total time which isn't half bad IMHO.

 

[bmcclure937]

Thanks for the reply. I am running 2x6TB WD Red drives with lower RPM for NAS use.

 

[Mr_N]

is it possible to run badblocks with just 1 or 2 passes over the drive instead of 4?

 

[nightshade00013]

is it possible to run badblocks with just 1 or 2 passes over the drive instead of 4?

that would be done with a -p X

as an example

badblocks -ws -p 2 /dev/adaX

This however will not be as thorough a test.

 

[Mr_N]

ahh cool thanks, yeah i know but they all passed long smart tests and they are quality drives so im sure 40 hrs instead of 80 will be ok.

 

[nightshade00013]

Ok, on the last two HGST drives I got I skipped a step and it will save a few hours if you want to do it this way. So I did a short smart test, no conveyance test since the drives do not support the test and then went straight to badblocks. After badblocks was done I did a short smart test followed by a long smart test. I figure it saved about 9 hours on the testing time by not doing a long smart test both before and after badblocks.

Since these were brand new drives I figured why not. The bad thing is that by skipping the first long test you could be potentially wasting a couple days on the badblocks test if the drive has a problem to begin with and with a used drive I would not likely skip the initial long test but with a brand new off the shelf drive I want to assume it is more than likely in good condition after 5 minutes of uptime and stress it then see if it passes.

It would still be better to do everything but if you are on a truncated time table and have new drives this should be a safe way to save a little time possibly leaving more time for a stress test. Your mileage may however vary and I am not responsible if skipping directly to badblocks eats all your data and your firstborn.

 

[Dave Martens]

This is a really great thread - thanks again to all for their contributions.

 

[Dice]

This might seem to be a slightly arcane/absurd question, granted I'm 38hrs into testing..
- At this moment I start to question if I really did the command "sysctl kern.geom.debugflags=0x10" before starting the tmuxes. I'm pretty sure I did, but yet again I fiddled around enough in tmux to get a feel for how the window/panes worked before getting into badblocks, that I've lost track on what was done or not. I hope I did the command. Is there a way to check? (or - does it really matter granted that the tests are indeed putting stress on the hdds?

Code:

Reading and comparing:  13.97% done, 38:24:15 elapsed. (0/0/0 errors)
96.88% done, 48:48:13 elapsed. (0/0/0 errors)

heh.

edit: aka re:reading the thread slightly less sleepy I found some answers in the thread. I'm just starting to get used to this method of "not trusting the info in OP - read all thread before doing anything" >.<
This post could safely be deleted >.<

 

[schofieldrj]

So I've started running these tests on my new drives, I haven't even finished the first long test and already getting some strange smart results on 3 of the drives...

Code:

  1 Raw_Read_Error_Rate  0x000f  100  100  006  Pre-fail  Always  -  3864
  3 Spin_Up_Time  0x0003  097  097  000  Pre-fail  Always  -  0
  4 Start_Stop_Count  0x0032  100  100  020  Old_age  Always  -  3
  5 Reallocated_Sector_Ct  0x0033  100  100  010  Pre-fail  Always  -  0
  7 Seek_Error_Rate  0x000f  100  253  030  Pre-fail  Always  -  1049
  9 Power_On_Hours  0x0032  100  100  000  Old_age  Always  -  5
 10 Spin_Retry_Count  0x0013  100  100  097  Pre-fail  Always  -  0
 12 Power_Cycle_Count  0x0032  100  100  020  Old_age  Always  -  3
184 End-to-End_Error  0x0032  100  100  099  Old_age  Always  -  0
187 Reported_Uncorrect  0x0032  100  100  000  Old_age  Always  -  0
188 Command_Timeout  0x0032  100  253  000  Old_age  Always  -  0
189 High_Fly_Writes  0x003a  100  100  000  Old_age  Always  -  0
190 Airflow_Temperature_Cel 0x0022  071  065  045  Old_age  Always  -  29 (Min/Max 24/35)
191 G-Sense_Error_Rate  0x0032  100  100  000  Old_age  Always  -  0
192 Power-Off_Retract_Count 0x0032  100  100  000  Old_age  Always  -  2
193 Load_Cycle_Count  0x0032  100  100  000  Old_age  Always  -  3
194 Temperature_Celsius  0x0022  029  040  000  Old_age  Always  -  29 (0 20 0 0 0)
197 Current_Pending_Sector  0x0012  100  100  000  Old_age  Always  -  0
198 Offline_Uncorrectable  0x0010  100  100  000  Old_age  Offline  -  0
199 UDMA_CRC_Error_Count  0x003e  200  200  000  Old_age  Always  -  0

Code:

  1 Raw_Read_Error_Rate  0x000f  100  100  006  Pre-fail  Always  -  9136
  3 Spin_Up_Time  0x0003  095  095  000  Pre-fail  Always  -  0
  4 Start_Stop_Count  0x0032  100  100  020  Old_age  Always  -  5
  5 Reallocated_Sector_Ct  0x0033  100  100  010  Pre-fail  Always  -  0
  7 Seek_Error_Rate  0x000f  100  253  030  Pre-fail  Always  -  901
  9 Power_On_Hours  0x0032  100  100  000  Old_age  Always  -  5
 10 Spin_Retry_Count  0x0013  100  100  097  Pre-fail  Always  -  0
 12 Power_Cycle_Count  0x0032  100  100  020  Old_age  Always  -  5
184 End-to-End_Error  0x0032  100  100  099  Old_age  Always  -  0
187 Reported_Uncorrect  0x0032  100  100  000  Old_age  Always  -  0
188 Command_Timeout  0x0032  100  253  000  Old_age  Always  -  0
189 High_Fly_Writes  0x003a  100  100  000  Old_age  Always  -  0
190 Airflow_Temperature_Cel 0x0022  077  075  045  Old_age  Always  -  23 (Min/Max 20/24)
191 G-Sense_Error_Rate  0x0032  100  100  000  Old_age  Always  -  0
192 Power-Off_Retract_Count 0x0032  100  100  000  Old_age  Always  -  5
193 Load_Cycle_Count  0x0032  100  100  000  Old_age  Always  -  5
194 Temperature_Celsius  0x0022  023  040  000  Old_age  Always  -  23 (0 18 0 0 0)
197 Current_Pending_Sector  0x0012  100  100  000  Old_age  Always  -  0
198 Offline_Uncorrectable  0x0010  100  100  000  Old_age  Offline  -  0
199 UDMA_CRC_Error_Count  0x003e  200  200  000  Old_age  Always  -  0

Code:

  1 Raw_Read_Error_Rate  0x000f  100  100  006  Pre-fail  Always  -  3832
  3 Spin_Up_Time  0x0003  095  095  000  Pre-fail  Always  -  0
  4 Start_Stop_Count  0x0032  100  100  020  Old_age  Always  -  5
  5 Reallocated_Sector_Ct  0x0033  100  100  010  Pre-fail  Always  -  0
  7 Seek_Error_Rate  0x000f  100  253  030  Pre-fail  Always  -  887
  9 Power_On_Hours  0x0032  100  100  000  Old_age  Always  -  5
 10 Spin_Retry_Count  0x0013  100  100  097  Pre-fail  Always  -  0
 12 Power_Cycle_Count  0x0032  100  100  020  Old_age  Always  -  5
184 End-to-End_Error  0x0032  100  100  099  Old_age  Always  -  0
187 Reported_Uncorrect  0x0032  100  100  000  Old_age  Always  -  0
188 Command_Timeout  0x0032  100  253  000  Old_age  Always  -  0
189 High_Fly_Writes  0x003a  100  100  000  Old_age  Always  -  0
190 Airflow_Temperature_Cel 0x0022  077  075  045  Old_age  Always  -  23 (Min/Max 20/24)
191 G-Sense_Error_Rate  0x0032  100  100  000  Old_age  Always  -  0
192 Power-Off_Retract_Count 0x0032  100  100  000  Old_age  Always  -  5
193 Load_Cycle_Count  0x0032  100  100  000  Old_age  Always  -  5
194 Temperature_Celsius  0x0022  023  040  000  Old_age  Always  -  23 (0 18 0 0 0)
197 Current_Pending_Sector  0x0012  100  100  000  Old_age  Always  -  0
198 Offline_Uncorrectable  0x0010  100  100  000  Old_age  Offline  -  0
199 UDMA_CRC_Error_Count  0x003e  200  200  000  Old_age  Always  -  0

The main thing giving me concern is the high raw_red_error_rate, these are brand new drives which have had no data on, just plugged in and began running the tests from the first post. Are these results something to worry about, should I look to return the drives or wait and finish all the tests first?

 

[BigDave]

I have a Seagate 6TB drive (see pic) and these drives report some things differently,
so much so that there are several posts in the forum regarding this.
This picture shows my SMART data for a drive that works fine, never had any issues with it so far.

 

[cyberjock]

In your case, it likely means something like too much vibrations on the disk, or something to that effect. Seagate has a totally different system for recording read error rates, so having some weird really large 6+ digit number is normal.

 

[Black Ninja]

I did follow the instruction and wanted to share my results. I have 16x4TB drives and ran the test on all 16 at the same time. It did worked and complete the tests on all of them without errors. It took 4 days aprox.

 

[cool77]

Hello, thanks to OP and all replies! It helps me doing these tests quicker!

Quick question that might have been already answered and I missed it.

While running badblocks test on a hdd, can we run a smartctl command, ie to get what are the actual HDD temperature so I can validate if my fans airflow ok.

Regards,

cool77

 

[nightshade00013]

Yes you can do a smartctl -a or smartctl -A to see what data is available just don't start doing tests while badblocks is running or you will end up having nothing but problems.

 

[cool77]

Thank you!

It now brings one more question. On my upper physical array, the upper hdds reach between 39-40C during badblocks. Lower ones about 36-38C. I use optimal fan setting with my x10sl7-f.

When using full speed fan setting during badblocks test, HDDs reach 31-32C on upper array and 29C on lower array.

Is badblocks worst load case on HDD? Or scrub? I just want to set the proper fan speed during production.

Regards,

cool77

 

[nightshade00013]

badblocks will likely be a worst case as it runs the drives very long and very hard.

 

[neto_hugo]

Guys... for how many patterns is recommmended running the badblocks test? Here I already have the 0xaa, 0x55, 0xff, 0x00 patterns without any erros (0/0/0). Does it would need more ???
My zpool is 5x6TB WD Red (RaidZ2)...

Thanks