Question Hard drive usage question

[jamesdsimone]

I have a number of NAS that I have alway used refurbed drives and have always had good luck with them. I have on average paying about 1 cent per GB. I have bought 8 Seagate 6TB EXOS drives and 10 Seagate 4TB ES.3 drives. The first 4 of the "refurbed" EXOS drives and 6 of the ES.3 drives were new so I just installed them without worrying about reliability. The second 4 EXOS drives were not new. They have the following usage. How do these numbers look? I've installed 1 and copied data to it.

1)817 days 13 power on cycles 13GB write/10GB read
2)818 days 9 power on cycles 90GB write/69GB read
3)995 days 8 power on cycles 111GB write/81GB read
4)1153 days 8 power on cycles 325 GB write/ 212GB read

 

[Tech Junky]

They may be 3-4 years old but, they don't look abused from a data standpoint. Should be good for a few years.

 

[jamesdsimone]

They may be 3-4 years old but, they don't look abused from a data standpoint. Should be good for a few years.

I got lucky with getting 10 out of 18 drives being new especially at less than a penny/GB. I'll install the other 3 EXOS drives in my NAS. The 4 used ES.3 have 2700+ days on them. I'll probably use them to archive an extra copy of my media collection.

 

[Tech Junky]

2700+ days

That would be a different story for me. That's pushing things a bit in terms of age. It's going to be the power cycles that kill them.

 

[jamesdsimone]

That would be a different story for me. That's pushing things a bit in terms of age. It's going to be the power cycles that kill them.

I just checked 2 of the ES.3 drives they both have 2730d 16hr but only a power on count of 3. They must have been installed, powered on and sat for that long before being pulled. I'm guessing they were both pulled from the same server at the same time probably for newer bigger drives. With that low of a power on count what do you think the reliability would be? I'm not getting read/write data like the EXOS drives. Is that because they are older? I'm using Speccy on both.

 

[Tech Junky]

Hard to say what they will do but a power count of 3 means the server didn't do much after being powered on. Probably a lab device that just didn't get used all that much. Probably just do an extended r/w test on them to make sure everything is in working condition.

 

[jamesdsimone]

Hard to say what they will do but a power count of 3 means the server didn't do much after being powered on. Probably a lab device that just didn't get used all that much. Probably just do an extended r/w test on them to make sure everything is in working condition.

I'll do that before deciding where to use them. I'll check the other 2 also. Suggestions on diagnostic software?

 

[Tech Junky]

You can use DD in Linux for free and once you start it just let it run until you stop it and then check with CDI for any issues.

 

[jamesdsimone]

I am having trouble with my latest build so no Linux box at the moment but hope to get it up and running soon.

 

[Tech Junky]

You might be able to use WSL if you can't do the USB boot. I haven't played around with WSL enough to test it.

 

[jamesdsimone]

You might be able to use WSL if you can't do the USB boot. I haven't played around with WSL enough to test it.

WSL? I downloaded Seatools since most of the drives I want to test are Seagates. I'll check it out and see if it meets my needs.

 

[Tech Junky]

Install WSL

Install Windows Subsystem for Linux with the command, wsl --install. Use a Bash terminal on your Windows machine run by your preferred Linux distribution - Ubuntu, Debian, SUSE, Kali, Fedora, Pengwin, Alpine, and more are available.

learn.microsoft.com

You could just run a ton of CDM tests as well from Windows.

 

[aigomorla]

8 Seagate 6TB EXOS drives and 10 Seagate 4TB ES.3 drives.

You were walking a tight rope the moment you went seagate.

I have a pair of EXOS in testing out, but they are the SAS versions.
They been holding out well, but your Connie drives the ES.3, i would not trust these with anything, and would consider swaping them out as they die to HGST's.

Infact i would probably have 2 on hand incase they do drop out, so you don't have long downtimes, and you can immediately resliver, and pray your redundancy holds out until you can finish the resilver.

hey must have been installed, powered on and sat for that long before being pulled.

You know some refurbished vendors wipe SMART data on the drive making it look new before they sell it, so they can do statics and see how long they last when they fail and you start a RMA.

https://www.reddit.com/r/DataHoarder/comments/105wtcx

So you really can't trust SMART data from a refurb / used drive.
The only time a SMART data will ever be true is if the BOX was sealed, and you got a factory sticker sealing that box, and you didn't buy them in bulk and got them in just a static bag.

A lot of us here don't like seagate, because we had too many fail on us.
I personally had lost i think 6 in one year, but they were the NAS drives, and this is before they moved to the ironwolf design.
But still they lost all trust with me, because when i moved to HGST, i been super happy with HGST He.
You can ask a lot of us who our favorite brand is, and a lot will say HGST He, or WD Gold, which is basically identical to a HGST He.

 

[jamesdsimone]

I have read that some vendors wipe the SMART data but why would they wipe the power on count and not the number of days? These drives are going into an NAS. Most of the data is mp4 files that are backed up on other drives. Losing a drive would be a pain but would not result in losing any data.

 

[aigomorla]

I have read that some vendors wipe the SMART data but why would they wipe the power on count and not the number of days? These drives are going into an NAS. Most of the data is mp4 files that are backed up on other drives. Losing a drive would be a pain but would not result in losing any data.

They like to do there own statics on what drive fails and after how many hours after recertification.
Its a common practice.
They don't lie to you up front saying its a new drive.
But its a common practice to wipe all smart DATA on some vendors.

 

[igor_kavinski]

You can ask a lot of us who our favorite brand is, and a lot will say HGST He, or WD Gold, which is basically identical to a HGST He.

I've heard that long term the helium leaks out and then the drive dies. What's the longest you have had a He drive run? Any sudden failures?

 

[aigomorla]

What's the longest you have had a He drive run? Any sudden failures?

I think i have a He06 running at the office for 6yrs.
I really can't see how the helium can leak unless your running your nas in some ganky environment where the seals and deteriorate.
Ive seen really bad cases of rust in PC's from running in Humid environments.

But if you take care of your NAS properly, i think the motor will fail before your Helium leaks out.

 

[jamesdsimone]

They like to do there own statics on what drive fails and after how many hours after recertification.
Its a common practice.
They don't lie to you up front saying its a new drive.
But its a common practice to wipe all smart DATA on some vendors.

They clearly stated they were refurbed drive so there is always a chance of failure. I'm replacing 2TB drives, 2 WD Red drives and 2 Seagate. I bought the enclosure in 2014 so the drives have been there almost 10 years.

 

[aigomorla]

Careful on the 2TB WD Reds.
They might be SMR.

On WD Red NAS Drives - Western Digital Corporate Blog

On WD Red NAS Drives - Data - Western Digital Corporate Blog

blog.westerndigital.com

 

[Micrornd]

I personally had lost i think 6 in one year, but they were the NAS drives, and this is before they moved to the ironwolf design.

Have you had problems with Ironwolfs' over 12tb ?

 

[aigomorla]

Have you had problems with Ironwolfs' over 12tb ?

no i havent went back to seagates until a friend of mine told me to try out the exos sas versions.
Those are the only ones i have from seagate, as i do not trust them as a vendor.