Other What are recommendations for installation of used HDD?

Y

YuryG

I can get a used HDD, somewhere it is recommended to "low format" it. What it is, is it possible in FreeBSD or is it beneficial? May be dd /dev/zero /dev/adaX or something to make all errors obvious to its controller?
 
Assuming it's been in service, just plug it in and collect the SMART report, as root, with something like smartctl –a /dev/ada0.

The FreeNAS Hard Drive Troubleshooting Guide will be invaluable in interpreting the results, i.e.

The important data we look at are as follows:

1) Serial Number
2) ID 5 Relocated Sector Count
3) ID 197 Current Pending Sector Count
4) ID 198 Offline Uncorrectable Sector Count
5) ID 199 UDMA CRC Error Count

Other notable data are:

6) ID 194 Temperature
7) ID 200 MultiZone Error Rate
8) Extended Self-test Time (value in minutes)
9) SMART Self-test logs, specifically the results of the self tests

​

 
Low format - use Victoria or MHDD.
Check for bad blocks there. If there are many bad blocks or the writing-reading speed every blocks is very high, the disk is not suitable.
 
I love to use used HDDs as can get some great bargains. Over the past 10+ years or so I have only encountered one failure and one dead/damaged on arrival.

If you look at specification sheets for drives they often estimate 1,000,000+ hours MTBF. Which is well over 100 years for our use. Granted, other factors do cause premature failure of drives, but overall they should be quite sturdy.

Depending on the model of drive there is also a conveyance test. I tend to run
smartctl -t conveyance /dev/daX

Note: I use a USB HDD adapter for that part.

If no errors from that I install into the machine and then use the
smartctl -t long /dev/adaX

As well, I tend to run in mirrored pairs so rebuild/resilver each drive singularly. This usually, not always, highlights errors if present.

A tip that has worked well over the years is to listen to the drive. Years and years ago you could diagnose slow HDD disk performance in Windows 3.11/'95 by listening to a drive. A significant noise difference was caused by the 32-bit driver not loading. If a drive takes excessively long to spin up or thrashes or ticks it will probably not be a healthy drive.

---

Droned on long enough....to answer question about low level formatting this might a hang over from the olden days when controllers used to be matched to disks. Best examples are on YouTube but:
View: https://youtu.be/uWWNbpd56Xs
is worth a look and modern drives don't need to "low level" formatted to match controllers these days.

---
Reading your question again maybe a distinction between low level and wiping would be useful.

To wipe I tend to use security/bcwipe which will wipe your drive to various options and confirm it has been wiped.
 
spanglefox said:
I love to use used HDDs as can get some great bargains. Over the past 10+ years or so I have only encountered one failure and one dead/damaged on arrival.

If you look at specification sheets for drives they often estimate 1,000,000+ hours MTBF. Which is well over 100 years for our use. Granted, other factors do cause premature failure of drives, but overall they should be quite sturdy.

Depending on the model of drive there is also a conveyance test. I tend to run
smartctl -t conveyance /dev/daX

Note: I use a USB HDD adapter for that part.

If no errors from that I install into the machine and then use the
smartctl -t long /dev/adaX

As well, I tend to run in mirrored pairs so rebuild/resilver each drive singularly. This usually, not always, highlights errors if present.

A tip that has worked well over the years is to listen to the drive. Years and years ago you could diagnose slow HDD disk performance in Windows 3.11/'95 by listening to a drive. A significant noise difference was caused by the 32-bit driver not loading. If a drive takes excessively long to spin up or thrashes or ticks it will probably not be a healthy drive.

---

Droned on long enough....to answer question about low level formatting this might a hang over from the olden days when controllers used to be matched to disks. Best examples are on YouTube but:
View: https://youtu.be/uWWNbpd56Xs
is worth a look and modern drives don't need to "low level" formatted to match controllers these days.

---
Reading your question again maybe a distinction between low level and wiping would be useful.

To wipe I tend to use security/bcwipe which will wipe your drive to various options and confirm it has been wiped.
Click to expand...
I don't think I need wiping.
 
gpw928 said:
Assuming it's been in service, just plug it in and collect the SMART report, as root, with something like smartctl –a /dev/ada0.

The FreeNAS Hard Drive Troubleshooting Guide will be invaluable in interpreting the results, i.e.

The important data we look at are as follows:

1) Serial Number
2) ID 5 Relocated Sector Count
3) ID 197 Current Pending Sector Count
4) ID 198 Offline Uncorrectable Sector Count
5) ID 199 UDMA CRC Error Count

Other notable data are:

6) ID 194 Temperature
7) ID 200 MultiZone Error Rate
8) Extended Self-test Time (value in minutes)
9) SMART Self-test logs, specifically the results of the self tests

​

Click to expand...
Then do smartctl -t long /dev/DISK_NAME. Wait for the specified time -- it will tell you. Then smartctl -a again. Toward the bottom of the report you will see a list of the last N tests. If any of the tests show as failed you're better off not using it. It'll save you a lot of grief.
 
spanglefox said:
I love to use used HDDs as can get some great bargains. Over the past 10+ years or so I have only encountered one failure and one dead/damaged on arrival.
Click to expand...
+42

I usually buy "salvage" disks (think: a business retires a multishelf SAN/DAS) for ~$1 -- regardless of capacity.

I'm not interested in anything smaller than a terabyte. And, if it has ANY bad/relocated sectors, it gets "thrown back".

I suspect I have well north of 200T, here... maybe even 300T. The issue becomes one of keeping track of "what is where"...
 
Useful information. As I've just encountered a drive that is failing, I think that is sound advice to record the SMART information at the time of first use so you have a data point. Perhaps a simple shell script is all that is needed and a git repository. The git repository would store the SMART stats of the drive and the script would help in data capture and format as well as comparison and perhaps suggesting a drive might need replaced.
 
You must log in or register to reply here.
Back
Top