I am using four12TB IronWolf drives in RAID-5 that have a recommended workload of 150TB/year. Just from the raid formatting (and adding of a 4th drive a few days after initializing) it was at 60TB workload a few days into existing.
If I do a bad block scan every month, that alone will hit the 150TB limits, and Segate's warranty page says they can deny claims if the workload is over the recommended value.
What do others do in this situation? I feel like a bad block scan is important to catch failing drives in advance, especially with large drives that might have data that isn't read for months until you need it. Theoretically a failing drive should start throwing catchable errors before it completely fails but I don't like trusting chance.
Also, is it normal for the drives to have 60TB of access recorded a few days after setup? That's reading/writing the entire drive 4 times which seems excessive, but it's been a long time since I used a RAID device and back then didn't even have workload stats.
How do most handle this? Run it twice a year maybe?
To monthly Bad Block Scan or not?
-
iansmith
- Posts: 12
- youtube meble na wymiar Warszawa
- Joined: Thu Dec 02, 2021 10:51 pm
-
SlimTech
- Posts: 9
- Joined: Mon Dec 06, 2021 8:20 pm
Re: To monthly Bad Block Scan or not?
Hi,
Personally I use 8TB WD Red Plus which offer 180TB/Y and WD Gold 4TB which offer 500TB/Y. I use the Reds to simply store stuff and the Gold to work with many reads/write since they should hold better over time and are faster.
I would say that the workload is just a placeholder to minimize the defective ones and minimize claims on them primary. Αt my work every Sunday is done a scrub and check on the servers, I have seen only 7 HDDs failing within weeks in 15 years and not covered under warranty. Most of the time they last at least 5 years before problems begin to appear. Personally I only had 2 HDDs fail on me.
I would say that you can do a check every month to verify the integrity of the raid and still not have problems for a minimum of 5 years.
Personally I use 8TB WD Red Plus which offer 180TB/Y and WD Gold 4TB which offer 500TB/Y. I use the Reds to simply store stuff and the Gold to work with many reads/write since they should hold better over time and are faster.
I would say that the workload is just a placeholder to minimize the defective ones and minimize claims on them primary. Αt my work every Sunday is done a scrub and check on the servers, I have seen only 7 HDDs failing within weeks in 15 years and not covered under warranty. Most of the time they last at least 5 years before problems begin to appear. Personally I only had 2 HDDs fail on me.
I would say that you can do a check every month to verify the integrity of the raid and still not have problems for a minimum of 5 years.