Hard Drive Burn-In Testing - Discussion Thread

[ShimadaRiku]

Yes. You get the benefit of verifying the condition of the drives and completely wiping them at the same time.

Was hoping to avoid a 3 day process of running badblocks

 

[Robert Trevellyan]

Was hoping to avoid a 3 day process of running badblocks

It's advisable, but not compulsory ;)

 

[pratman2]

Can anyone elaborate on this statement in the tutorial?

"Now, before we can perform raw disk I/O, we need to enable the kernel geometry debug flags. This carries some inherent risk, and should probably not be done on a production system."

Aside from the risk of punching in the wrong disc number and corrupting an existing volume when running badblocks, what is the level of risk? Is it possible to loose data on an existing volume even if running badblocks on the correct disc?

 

[pratman2]

Sorry, should have read your post better. Are you logged in as Root? If it can't find ada0, then are you sure that's the name of the disc? Could it be da0?

Yes I logged in as root as well and same issue. Using ada0 for the drives it would let me run the SMART tests but when I tried to enable "kernel geometry debug flags" it would give me that error and not let me go any further.

 

[okgunguy]

Yes I logged in as root as well and same issue. Using ada0 for the drives it would let me run the SMART tests but when I tried to enable "kernel geometry debug flags" it would give me that error and not let me go any further.

I just ran this procedure last night on 2 extra 1Tb drives. But I didn't pay attention after the kernel command. I think it gave me the same error, but it still let me run tmux and badblocks. Is yours locking up and not giving you a cursor back? If so, that is beyond my knowledge, something wrong is happening there. If it gives you your cursor back, then just run tmux and badblocks. They'll run (forever) and then you can run your last SMART long test and check your results. If everything is still 0's and haven't incremented, then you're good to go.

 

[cyberjock]

Can anyone elaborate on this statement in the tutorial?

"Now, before we can perform raw disk I/O, we need to enable the kernel geometry debug flags. This carries some inherent risk, and should probably not be done on a production system."

Aside from the risk of punching in the wrong disc number and corrupting an existing volume when running badblocks, what is the level of risk? Is it possible to loose data on an existing volume even if running badblocks on the correct disc?

You can also have kernel panics and whatnot.

 

[pratman2]

I just ran this procedure last night on 2 extra 1Tb drives. But I didn't pay attention after the kernel command. I think it gave me the same error, but it still let me run tmux and badblocks. Is yours locking up and not giving you a cursor back? If so, that is beyond my knowledge, something wrong is happening there. If it gives you your cursor back, then just run tmux and badblocks. They'll run (forever) and then you can run your last SMART long test and check your results. If everything is still 0's and haven't incremented, then you're good to go.

Logged in as root, still not taking the kernel command, but I moved on to temux and starting badblocks, it has worked fine. Badblocks test on the drives is almost finished. Thanks for your help.

 

[Fuganater]

So I have badblocks running on 17x 2TB drives. I am trying to see the status but "tmux attach" doesn't really do anything for me. What should I be typing in to see either all drives progress or just drive?

EDIT: Did some more googling and figured it out. My only issue now is when i am in tmux, I have no idea how to get out of it. Currently I just close Putty and relaunch....

 

[solarisguy]

[...] EDIT: Did some more googling and figured it out. My only issue now is when i am in tmux, I have no idea how to get out of it. Currently I just close Putty and relaunch....

CTRL-b d
That is press CTRL-b, then press d

 

[Fuganater]

CTRL-b d
That is press CTRL-b, then press d

Thanks! I need to make my own cheat sheet.

 

[Fuganater]

So badblocks has been running for over 2 days now... looks like it is running the same test over and over. Is that right? Can I say that this ran ok and I should stop it now?

 

[Ericloewe]

So badblocks has been running for over 2 days now... looks like it is running the same test over and over. Is that right? Can I say that this ran ok and I should stop it now?
View attachment 9179

It's normal. It's testing different patterns.

 

[Fuganater]

It's normal. It's testing different patterns.

How many patterns will it go through?

EDIT: Found the wiki on badblocks. Looks like default goes through 4 patterns. I am on pattern 3/4 so I probably have another 36 hours left.

read-write Test (warning:destructive)
This test is primarily for testing new drives and is a read-write test. As the pattern is written to every accessible block the device effectively gets wiped. Default is an extensive test with four passes using four different patterns: 0xaa (10101010), 0x55 (01010101), 0xff (11111111) and 0x00 (00000000). For some devices this will take a couple of days to complete.

 

[STREBLO]

Man, this is killing me, I'm so impatient.

 

[Fuganater]

Haha I know the feeling. At least you are on the last run. Mine completed at about 96 hours.

 

[Ericloewe]

Man, this is killing me, I'm so impatient.

Meh, that's nothing. I ran three runs of badblocks. I only had room for four drives at the time, and I didn't want to tickle my OCD by testing without four drives in the bay and/or unevenly testing the drives.

 

[STREBLO]

Haha I know the feeling. At least you are on the last run. Mine completed at about 96 hours.

Second last :( At 90 hours now, still have the last pass to go.

Edit: Of nevermind, you were right. Only 15% on my last disk to go!

Meh, that's nothing. I ran three runs of badblocks. I only had room for four drives at the time, and I didn't want to tickle my OCD by testing without four drives in the bay and/or unevenly testing the drives.

Why run more than once, wouldn't it just check the same thing?

 

[rogerh]

Second last :( At 90 hours now, still have the last pass to go.


Why run more than once, wouldn't it just check the same thing?

What he said - you don't want to treat any of the disks differently from the others - takes 3 passes!

 

[Ericloewe]

What he said - you don't want to treat any of the disks differently from the others - takes 3 passes!

Besides, the rest of the server hadn't arrived yet. :p

 

[STREBLO]

What he said - you don't want to treat any of the disks differently from the others - takes 3 passes!

Oh of course, i'm letting them all finish. I thought he meant running the entire badblocks through to completion three times.