Have five 2tb drives w/over 8yrs power on time, what do you think about...

[DaveBetTech]

.... overall reliability?  HERE'S THE SITUATION:  I had a MS storage space w/ReFs and integrity streams enabled. I noticed that my space was screwed up showing 100% utilization vs actual 62%. Since I had redundancy, I pulled one of the drives out and moved the data over to a 10tb drive, recovering the data.

 

Well, during the move, some of the files FAILED the CRC check from the integrity stream, about a dozen files out of maybe . Well, shit.

 

DRIVE HISTORY OVER LAST ~10 YRS: I've had drive sleeping disabled. The server stayed up most of the time, so only 60 power-on counts, or about 1 power-on every 2 months roughly. I have a decent PSU and connected to a good surge protector. There are rubber stoppers connecting the drives in their bays too. Temperatures are not great even with airflow, usually around high 20s, low 30s. USAGE: These drives are mostly used for a Google Drive backup and Plex Server storage.

 

WUDDA-YA THINK I SHOULD DO ABOUT THE 2TB DRIVES?: The SMART data doesn't show anything bad for all five 2tb drives. I just did a chkdsk /r /x /f and it came back 100% clean from these freshly formatted drives (took about 5 hrs each), zero bad sectors. Do you think

1. I should dump these 2tb drives and replace with a 10tb, do a 2 way mirror w/ReFs & integrity streams like I was doing before?
2. I should keep these 2tb drives, add that extra 10tb and maybe do a THREE way mirror w/ReFs & integrity streams for added reliability? 
3. I should keep these 2tb drives, add that extra 10tb and do a 2 way mirror w/ReFs & integrity streams, adding 5tb of effective space (don't really need it tho)?
4. Some other option I haven't thought about

 

I'm not interested in moving to Linux atm. I don't know enough about it if shit hits the fan. Also, I want to keep hardware transcoding support w/my 1060 in Windows (unless they've added Linux support).

 

ALL DRIVES HAVE SAME CHKDSK RESULT (checking 10tb drives later):

 

SMART DRIVE OUTPUT FOR ALL 2TB OLD-ASS HDDs:

[Electronics Wizardy]

Id just keep the current 2 way mirror. Just restore backup is something bad happends to multiple drives at once.

 

You can replcae the drives, its just how much do you want to spend. THose drvives are pretty heavily used though.

[DaveBetTech]

@Electronics Wizardy Thank you for replying. I don't know what caused the file issues beforehand and just assumed it was the old-ass hard drives. I haven't done a full chkdsk scan on the two 10tb drives I have, but the 2tb came out clean and no SMART issues.

 

I don't have a backup solution either, just these drives. I do backup all my personal data on Google Drive (photos, important personal docs, etc) and this is stored on the Storage Space as well. I use Google's Backup And Sync app to keep it constantly backed up.

 

I've already paid $270 to replace the five x 2tb drives with a single 10tb Seagate Ironwolf, just waiting on the shipment. I'm just considering if I should trust these drives for the next two years or just three 10 tb drives.

[Electronics Wizardy]

Just now, DaveBetTech said:

@Electronics Wizardy Thank you for replying. I don't know what caused the file issues beforehand and just assumed it was the old-ass hard drives. I haven't done a full chkdsk scan on the two 10tb drives I have, but the 2tb came out clean and no SMART issues.

 

I don't have a backup solution either, just these drives. I do backup all my personal data on Google Drive (photos, important personal docs, etc) and this is stored on the Storage Space as well. I use Google's Backup And Sync app to keep it constantly backed up.

 

I've already paid $270 to replace the five x 2tb drives with a single 10tb Seagate Ironwolf, just waiting on the shipment. I'm just considering if I should trust these drives for the next two years or just three 10 tb drives.

The file intergrity steams should make it so the drives can't cause bad data to be read or written, so it could be something else.

 

Id make that 10tb drive a backup for this array so a array issue won't cause data loss, raid array can fail as a whole.

[DaveBetTech]

@Electronics Wizardy Damn, that's a good idea that I never contemplated. I'm going to strongly consider this.

[DaveBetTech]

I found an article where someone ran a full chkdsk & found no problems but WD's Data LifeGuard Diagnostics Extended Test DID find errors.

 

I'm going to use this and test all five of my 2 tb drives. I might feel better after that, we'll see.

[DaveBetTech]

Well, with full scans from chkdsk and Western Digital tools, it found zero problems with all 2 tb drives and the two 10 tb drives. I ordered a 10 tb extra drive and decommissioned two of the 2 tb drives. 

 

I setup another storage space w/ReFs file system, enabled integrity streams and disabled enforcement. After doing a full scan of the extra 10 tb IronWolf, I'll move the ~8 tb of data to the storage space. I should have about 16.3 Tbytes of effective storage, which should last a super long time.

 

Since I live in an area with periodic brownouts due to frequent storms, I am finally buying a UPS. I'm also grabbing a 200mm fan to help push more air near the HDDs to keep them all below 30°C. Figured this will really help with overall stability because it'll never experience another power outage or brown out, not to mention with line interactive AVR, cleaner power.

 

Well, that's the idea anyways. Thanks for the advice.

[DaveBetTech]

BTW: I used StableBit Scanner to determine that most of my drives were well beyond designed limits, so I just got rid of all my 2tb old drives. It also significantly lowered wattage too, as my UPS shows a much higher uptime after those five drives were replaced with a single 10tb. Oh, and, it also told me that my NVMe 240gb game drive is hosed. I am in the process of RMA'ing it.

 

Also, note that StableBit Scanner doesn't recognize individual drives once you put them into a Storage Spaces array.