Disk errors - are these really critical?

[joeschmuck]

I remember needing to run Seagate tools for an RMA. They wanted some code back in the day. Remember, 30+ years ago all consumers were incompetent according to tech support. I like WD support, like @Ericloewe said, you tell them it fails SMART and you get an RMA. Okay, it's an online form the last time I did an RMA and I did an advanced replacement so it was just easier overall.

 

[danb35]

Remember, 30+ years ago all consumers were incompetent according to tech support.

...and that is no less true today--the large majority are incompetent. And many of those don't even know it, unfortunately (i.e., the Dunning-Kruger effect). So when a tech support rep assumes the caller is incompetent, he's on pretty safe ground. The problem, of course, is that some customers are quite competent, and in some cases rather more so than the tech support reps.

 

[joeschmuck]

The problem, of course, is that some customers are quite competent, and in some cases rather more so than the tech support reps.

LOL, yes so true. Today I got a call from my company's IT folks to try and fix a problem I had (all my 2011 emails lost on the company server, I suspect someone screwed up a retention policy). They look to see if I accidentally deleted the email and concluded I didn't. They started talking about some other crap and I started to correct them. While I do not claim to know much about how our email works, I did realize they didn't either.

 

[Dice]

While I do not claim to know much about how our email works, I did realize they didn't either.

From my experience the only way to get something useful out of the IT department 1st line is to come prepared with a detailed description on how to fix the problem at hand.
Even if it cost half a days worth of work, it would save weeks of frustration to get it done out the gate.

 

[Robert Trevellyan]

Seatools is present on the Ultimate Boot CD. You can shut down your FreeNAS, disconnect all but the flaky drive, boot from UBCD and do the necessary.

 

[rs225]

It may seem pointless, but anybody who knows, knows that just making the customer disconnect and reconnect everything makes a bunch of problems magically disappear... because they weren't product defect problems in the first place. I realize this is hypothetical, because it seems pretty clear that the OP drive is not healthy.

In other words, "Check both ends of the cable for damage on the connector" is codeword for "make sure you have it plugged in."

 

[Z300M]

It may seem pointless, but anybody who knows, knows that just making the customer disconnect and reconnect everything makes a bunch of problems magically disappear... because they weren't product defect problems in the first place. I realize this is hypothetical, because it seems pretty clear that the OP drive is not healthy.

In other words, "Check both ends of the cable for damage on the connector" is codeword for "make sure you have it plugged in."

On one occasion, my ISP (Charter) had me unplug the cable between the cable modem and the router and connect it the other way round! It still didn't work, because the gas company had dug up the Charter cable down the street, but Charter said they had had no reports of an outage.

 

[ChiknNutz]

Well I went ahead and did as they asked. I hunted for a suitable ISO creator (Rufus refused to load the Seatools ISO) and eventually created a bootable USB. I then had to round up another monitor, keyboard and mouse. I connected those up to the NAS and then started the Seatools application. I am currently in the process of running a Long Test, though it is far from complete. I started the test just under a little over two hours prior to this posting, and it is only 0.38% complete! If this is any indication of how long it will take, it could take days to complete.

Oh, I did receive another email from Seagate after I replied with, "Do you really, really, really need me to do all this?" and they said they would honor the RMA. BUT...the email also states that they will perform a test upon receipt and may still opt to deny the warranty claim. Nice.

 

[tvsjr]

BUT...the email also states that they will perform a test upon receipt and may still opt to deny the warranty claim. Nice.

That's a nice hard drive you have there... sure would be a shame if something "happened" to it (like it falling a few feet and landing hard on a desk while spun up).

Purely hypothetical, of course.

 

[ChiknNutz]

That's a nice hard drive you have there... sure would be a shame if something "happened" to it (like it falling a few feet and landing hard on a desk while spun up).

Purely hypothetical, of course.

LOL.

Well the Long Test did complete and the result is as expected...it failed, so there should be no reason they won't accept it now. The Seatools app suggested performing a zero based erase, which I have initiated. The claim is that it can sometimes resurrect otherwise bad drives. I guess we will see.

 

[Z300M]

That's a nice hard drive you have there... sure would be a shame if something "happened" to it (like it falling a few feet and landing hard on a desk while spun up).

Purely hypothetical, of course.

Do Seagate drives contain shock sensors, whose state would be detected by SeaTools but would not show up in a SMART test?

 

[ChiknNutz]

QUESTION...During this initial RMA process, Seagate is saying that they may replace the drive with one of larger capacity. I know I cannot take advantage of any increase in size after initially creating the pool, but is it safe to replace a failing drive with one of larger capacity? I recall reading about this in the beginning when I set this up a couple years ago, but to be honest, I have simply forgotten the protocol as related to this issue.

 

[tvsjr]

Yes. Down the road, you could replace the other drives with larger models (one at a time) and gain extra space on the vdev/pool.

 

[Stux]

By default, FreeNAS configures the pools to auto-expand to the largest common disk size.

If auto-expand is disabled on your pool, you can manually enable it, or manually force the expansion, once all the disks are larger.

 

[ChiknNutz]

Thanks for the help so far. So in the time it took to get the replacement from Seagate (arrived today), another drive appears to have failed! I am not quite sure how to determine which disk to replace with which since now what I am presented with in the "volume status" are two disks that are UNAVAIL and all they show (rather than their previous name) is a long string of unintelligible numbers. Do I just pick one and hope for the best?

 

[Reece Johnson]

I'm not the most knowledgeable source on FreeNAS, but I'll offer what little help I can.

First question: what is the status of your backups?

From your signature you are running RaidZ2, now with two failed drives. If you care about your data at all, I would check your backups before touching anything. One bad block anywhere on the remaining four drives could trash your pool.

Next, I would correlate the working serial numbers with the drives in the chassis. Make sure these working drives are not removed.

 

[ChiknNutz]

I perform a backup of important data (not the whole volume) daily. I know which two are the offenders, but the display in the FreeNAS GUI is not helpful with the two offline disks. I have replaced one and the resilver process is almost complete.