Saturday, January 4th., 17:32 CET, an e-mail arrived in my Inbox, sent from my TrueNAS server:
New alerts:
* Pool volume0 state is DEGRADED: One or more devices has been removed by the
administrator. Sufficient replicas exist for the pool to continue functioning
in a degraded state.
The following devices are not healthy:
* Disk TOSHIBA HDWN160 10M3K1CNFAXG is REMOVEDNo warning, whatsoever. Double checked my e-mails, looked in the TrueNAS alerts. Nothing. Apparently the disk just disappared out of the blue. Or is it into the dark, when stuff disappear? 🤔
Of course I had a cold standby disk ready in a drawer and after consulting the TrueNAS documentation on how to replace a failed disk, I had the NAS back online and resilvering within 30-40 minutes. That I had to consult the documentation also shows, that it’s not that often I have to do this task.
But this is actually different. When I made this DIY NAS back in 2018, I chose to purchase different brands of disks over some months, both because eight 6 TB disks aren’t cheap and to hopefully spread them over multiple batches.
The original disk list was:
- 2 x Seagate IronWolf ST6000VN0033
- 1 x Western Digital Red WD60EFRX
- 1 x Toshiba N300 HDWN160
- 4 x Seagate Barracuda ST6000DM003
Four NAS grade disks and four consumer grade disks. The Barracuda 6TB’s has been a huge disappoinment. I had looked at the Backblaze Hard Drive Data and Stats and found that the Barracudas weren’t much worse than dedicated NAS disks, but they didn’t have any 6 TB disks, only 4 TB and 8 TB.
None of the original four Barracudas remain. Two failed within warranty and was RMA’ed. One is actually still in the NAS with 11 145 Power_On_Hours (as per smartctl). It replaced another Barracuda (at 19605 Power_On_Hours) that started to get Offline_Uncorrectable errors. These errors were resolved when running a ‘badblocks’ test on the disk and it is now serving in my offsite backup QNAP TS-419P NAS.
For comparisson, the IronWolfs have 56 939 power on hours, which equals six years and 160 days.
But this was the first NAS grade disk to fail and I have three more of the same model (with 48 443, 52 578 and 56 303 power on hours respectively), because it has been the cheapest 6 TB disk on the market for quite some years now and when a Barracuda failed it was either replaced with an RMA’ed Barracuda or a Toshiba N300.
The new cold standby disk will be a Seagate IronWolf 6TB, which is already in the mail, according to the webshop.
Before discarding the Toshiba disk completely I will perform a ‘badblocks’ test on it, if it can be detected at all on my test computer.
If the disk is unrecoverable, I will hand it in to my local recycling station, after removing the circuit board which is held by 9 small philips screws. If someone finds the disk and happens to have a similar circuit board, it still only contains an eighth of the RaidZ2 data.