Hi,
I would like to clarify whether the following behavior is expected on QuTS hero, or if this points to a software or hardware issue.
On a freshly initialized QNAP TVS-h674 running QuTS hero (ZFS), the following behavior is observed:
• HDDs repeatedly spin down and spin up, even when idle
• Audible noise from frequent power state changes
• Sequential reads (e.g. MKV playback, large archive files) occasionally stutter
• S.M.A.R.T. attribute ID 4 (Start_Stop_Count) increases rapidly
What has already been verified:
• S.M.A.R.T. polling set to 60 minutes
• No scheduled S.M.A.R.T. tests (short or long)
• No snapshots
• No pool scrubbing / background jobs
• No deduplication
• HDD standby explicitly set to Never
• System pool / metadata on SSD also tested (no change)
Despite this, disks still enter standby and are reactivated by the system.
QNAP support stated that system-initiated disk wake-ups cannot be fully eliminated on QuTS hero and suggested switching to QTS (EXT4) if this behavior is undesired.
System information:
• NAS model: QNAP TVS-h674-i5-32G (4711103082386)
• Version: QuTS hero h5.2.8.3359
• Expansion: QNAP QM2-2P-344A PCIe card installed for 22110 (110 mm) NVMe SSDs
• Datapool 1 (RAID 1): 2× SSD SAMSUNG NVMe-SSD PM9A3 960 GB M.2 (MZ1L2960HCJR-00A07)
• Datapool 2 (RAID 6): 6× HDD WD Gold 12 TB (WD122KRYZ)
Workload:
• Archive data
• Media files (MKV)
• Backups
• No databases, VMs, containers, or heavy random I/O
Questions:
- Is frequent HDD spin down/up considered normal behavior on QuTS hero?
- Is stuttering during simple sequential reads expected with ZFS on QNAP?
- Has anyone observed similar behavior on comparable systems?
- Would switching to QTS (no hero) be a valid way to rule out hardware issues?
Thank you for any technical insight