Cheap 18TB Seagate Drives... ebay.us/tBC10b (ServerPartsDeals) Digital Spaceport Drives... shop.digitalspaceport.com/ Pico Power Supply... amzn.to/3L987H0 (affiliate links, support this channel!)
I have not seen anyone measure power usage accurately before and found the results interesting as I have been thinking of a way to measure power use of the disks. I use ZFS Raid Z2 for the plot storage on my small 27TB farm. I use 6TB drives which have been redirected from the bin because they have a few bad sectors, so I can offset the zero cost of disks against the higher power cost. Occasionally during plotting a drive was kicked out but could be added back into the raid, and if a bad sector is discovered during farming, ZFS returns the correct data to the farmer and the error count increases. Also I don't have a machine capable of plotting at any speed so losing a single disk of plots would take significant time to re-plot, but you have powerful plot machines and am curious how many KWHs it would cost to re-plot the contents of a failed disk. I completely agree that single disks work for your setup and wanted to give the view of a small farmer and why I chose raid.
I think an overall power draw would have been better, including fans and other normally connected devices. if you wanted to get the actual impact on power used
I do agree that would have been a measurement more reflective of real-world useage; however, it's difficult to get high resolution readings that way and we're talking about a difference of a few watts.
@@HomeSysAdmin I guess you are getting accurate representational results. I imagine that as you use the hard drives, they would produce more heat in themselves, and the fans of the motherboard and the CPU and power supply would draw a bit more
Pretty cool experimentation. Not surprised by the results, pretty sure because of raid configuration, none of these HDDs can go in idle mode where the main power saving happens. Have a great week ahead and keep it up the great job 👍🏽💪🏽
wait did you just say you wanted to get only the disk drive level of power used.. a raid has alot of processing back at raid cards and cpu to maintain the raid.. versus a plain hard drive write and done.
Yes, as close a I can. I'm aware it's not perfect - such as the losses in the power supply aren't considered here. I used a software array (md) and it's level 0 (basic striping ith no parity) so while you are correct that there would be some additional power considerations for the CPU overhead, I'm not sure it would be much in this scenario.
@@HomeSysAdmin the difference is Power modes. RAID0 with Chia IO Profile means all HDD's never enter Idle1 Mode. Idle1 is around 10-15% lower in Power consumtion. I have do simulare Tests, But i have only Check Power Mode Timer to get a idear what Setup suites me better. Power Draw are documented in Spec Sheet.
