Load cycle count on IronWolf

[revengineer]

Hello,
I am starting to buy drives for my freenas. I was very uncomfortable with Western Digital shadiness on SMR drive. While there is a current list of those drives published now, it left a bitter taste with me. So after 15 years I switched from Western Digital to Seagate Drives. Time will tell whether the move was good or bad, but I do have a questions on the load cycle count: On WD drives I ran the WDIDLE3 utility to disable the head parking to reduce the load cycle count, which on some drives grew excessive over time. Can this be done on Seagate IronWolf drives as well? If so, what utility do I need an how do I use it? Advice appreciated.

Thank you!

 

[danb35]

It isn't necessary with the Ironwolf drives. I have a couple in my system; one example, in 32k power-on hours, has a load cycle count of 23k. Higher than it would be on a WDIDLE'd WD, but well within limits.

 

[revengineer]

Thank you @danb35. Follow up please: Should I be setting the APM to >128 to prevent spin down? What does the drive do different for setting 128 vs 254 to save power?

 

[danb35]

I'm honestly not sure what that setting does; I've never changed it on my system.

 

[revengineer]

Ok thank you. I will do some more reading but will not worry for now. Still working on the burn in. These are 8 TB drives and it will take a while, so there is no rush. :)

 

[gary_1]

Bumping this as I've just bought a 4TB Seagate Ironwolf to replace a failing 4TB WD Red (similar reasons to OP for moving away from WD). I noticed after just 19 hours run time that the load_cycle_count is at 160 already and whilst typing this post it's increased to 165. Do neither of you see the same rate with your Ironwolf?

Specs for the drive suggest a 600,000 load cycle count limit, so it'd take some years to reach that limit, but even so seems odd for a NAS aimed drive to be doing this as a default?

Also, on the APM page, what's the difference between standby and spindown? I assume spindown is what is causing the increasing load cycle counts, but then what would "standby" do?

I guess given it's 165 load cycles over 19 hours, that's about 208 per day it woud take just about 8 years to hit the spec. Although the other concern aside from lifespan, is, if it's spinning down, won't that impact the pool performance/latency if the other 5 drives (WD Red) are not?

 

[Hellione]

Here as a example, first two are Exos, third is a Ironwolf, which i reused in a newly built system:
power_on_hours 589 load_cycle_count 3598 (was a new drive, start_stop_count 3)
power_on_hours 589 load_cycle_count 3604 (was a new drive, start_stop_count 3)
power_on_hours 995 load_cycle_count 4473 (was used as a backup drive before, last 589 hours in new truenas, start stop count 174)
seems to be about a 1:7 ratio ^^

 

[ChrisRJ]

As to the Exos drives, IIRC they have a two minutes threshold for this. Since they are data center drives, where idleness of more than two minutes is a very rare condition, on a low-load private NAS that could happen. I have the Exos drives myself, and changed the threshold to five minutes. There was a recent thread where @Samuel Tai provided the command to change that threshold.

 

[Hellione]

ok thank you, i will search for this thread, just to know how to change :)
otherwise, no need to do anything, i calculated about 52500 per year, so over 10 years to reach the threshold.
and i can´t remember to use any drive longer than 6-7 years