Critical Error - Sector unreadable

[abe_one]

Hello to all,
I write here because the Italian forum is poorly attended.
Saturday I received an email from freenas telling me: "critical errors."
the alert message in the web interface was: Device: / dev / ADA3, 2 Currently unreadable (pending) sectors (changed +1).

I tried to give a restart, it got worse until the status of the raid became: DEGRADED.
I connected the monitor to the server, and I saw it was started some process, after which the raid was returned: HEALTY, but the number of unreadable sectors ADA3 unit has grown to about 70.

I have not yet launched any of the smart status check, no nothing. After the long process I managed to make a back up from scratch of sensitive files. Remain about 15 TB of movies that I can not save elsewhere.

I looked at some official guides and video about replacing a failed disk, and they show the procedure, where the disc (in the resulting list ONLINE - while all my records are ONLINE), clicking on "replace" and select the parity disk it is not in the list.
Also I have a raidz1 but in the list of disks that make up the volume appear to me all 6 discs which then make up the replace I can not do.

I had thought to extract the ADA3 disc, initialize it to another computer and put it back like new, then in theory the system should rebuild the new drive without unreadable sectors, but I'm afraid it is a bullshit.

I await your advice because I'm a bit anxious about how to move to avoid worsening the situation.

Thanks for your help and sorry for google translate.

 

[hugovsky]

Just posting your hardware so everybody can see it:

MB: asRock e3c226d2i
CPU: xeon E3-1230v3
RAM: 16gb ECC
OS: Dual Cruzer Fit USB3 16gb
HDD: 6x4tb WD red Z1
COOLING: Noctua (3x Fan + CPU)
CASE: Silverstone DS380 8x3,5 Hot swap
PSU: Silverstone ST45 80 Plus Gold SFX
UPS: APC Smart Ups Pro 1500

RAIDZ1 with your size of disks is not a very good move. You should change it. You must replace your failing drive ASAP and expect to lose your pool while resilvering. Well... Maybe not, but I'm giving you the worst case scenario so you can be prepared.

Your disk is dead. Shut down the server, replace the failing drive and hit replace in GUI afterwards. It should start resilvering and when it ends, your pool should be ok. Did I mentioned RAIDZ1 is dead and not an option with 1GB+ drives?

 

[gpsguy]

Don't try "fixing" the failing disk. Get a NEW hard disk to replace the bad one.

If you have space for another drive and a spare SATA port, you could do the replace/resilver, with the old (bad) drive still connected.

The parity that you have with RAIDz1 is not on a single disk, but spread across all the disks.

 

[abe_one]

Thank you all.

I did not understand very well what you said about the resilver. That is the party and concluded without giving indeed problems ....


What is the procedure for disk replacement?
From off system replaced the defective disc - turn - aside ricostruizione automatically?
I looked at the manual but I did not understand, speaks of the replace button but I do not in my case.

I have no additional SATA ports. I had planned to take a PCI card with 2 or 4 sata.

I await advice on how to continue, in the meantime contact the vendor to have a change in the guarantee. But if I do this procedure the disc before I have to deliver and it must be tested to verify the defect, I of that record I can help it, or do I need?

Thank you

 

[rs225]

1st: You need to buy a new drive.

Do not contact the warranty/guarantee until after the problem is finished.

2nd: Have you done a scrub?

Maybe it automatically scrubs? Before you start the replacement, you should be sure a scrub as been done in the past 30 days. Attention: The scrub may cause ada3 to show more errors, or worse errors. This is okay.

3rd: Replace ada3 with the new drive.

Identify which disk is ada3. You must not make a mistake. Follow the direction in the manual.