October 28, 20187 yr When downloading via NZBGet the system becomes nearly unusable. The specs are: - Ryzen 2700x - 32 GB DDR4 Corsair LP - Crosshair VII Hero The array is xfs (encrypted): - Parity: 8TB WD RED NAS - Disk 1: 4TB Seagate - Disk 2: 4TB Seagate - Disk 3: 8TB WD RED NAS The cache is: - Kingston SSDNow V3 120GB (Will be replacing with 1TB Samsung 970 EVO NVME in 2 days) All downloads from NZBGet go directly to the SSD cache (cache option for the share is set to "yes"). I've also attempted to pin NZBGet to the last 4 core & threads with no luck. Netdata is reporting extremely high I/O wait times as seen below: htop is reporting extremely low CPU usage but the dashboard is reporting 100% usage: I've got TRIM (dynamix plugin) scheduled to run hourly. A test using "DiskSpeed" reported that the cache drive should be operating normally: I'm not running any VMs - just docker containers. Would anyone have any suggestions where to begin to troubleshoot this issue? Please let me know if more information is required. I do believe that this hardware should be more than capable? Edited October 28, 20187 yr by fletchto99
April 24, 20197 yr I got this working by buying a second SSD and putting my downloads on that. TBH, it was the single best thing I did after I got unRAID setup. I am in the process of building a second box now and will definitely have two SSDs. I know it sounds insane and nagged at me (I wasn't able to figure it out) but $40 down for a 512GB was worth it. Everything is great now.
April 24, 20197 yr Yeah, mine is only at sata2 speeds so SSD not sure how much benefit it will do. I may upgrade either way.
Archived
This topic is now archived and is closed to further replies.