[yurishimo]

I don’t think anyone would care if Synology gave priority to their own drives. A checkbox during setup that says “Yes, I know I’m using these drives that have not been validated blah blah blah” would be plenty. That’s not what Synology did however and that’s the main reason everyone is pissed.

[netruk44]

Funny you mention that…

I have a DS1515+ which has an SSD cache function that uses a whitelisted set of known good drives that function well.

If you plug in a non whitelisted ssd and try to use it as a cache, it pops up a stern warning about potential data loss due to unsupported drives with a checkbox to acknowledge that you’re okay with the risk.

So…there’s really no excuse why they couldn’t have done this for regular drives.

[spease]

That assumes that the person setting up the NAS is the same person using it, which is not going to be the case for non-tech-savvy users.

Everyone will understand it costing more, fewer people will understand why the NAS ate their data without the warning it was supposed to provide, because cheap drives that didn’t support certain metrics were used.

If Synology wants to have there be only one way that the device behaves, they have to put constraints on the hardware.

[brookst]

But those people would call support when the array couldn’t rebuild. And many of them would blame Synology, and demand warranty replacement of the “defective” device, and generally cost money and stress.

As long as Synology is up front in the requirement and has a return policy for users who buy one and are surprised, I think they’re well within their rights to decide they’re tired of dealing with support costs from misbehaving drives.

As long as they don’t retroactively enforce this policy on older devices I don’t understand the emotionality here. Haven’t you ever found yourself stuck supporting software / systems / etc that customers were trying to cheap out on, making their problems yours?

[kstrauser]

It’s not that I don't understand. It’s that as an end user I don't give a shit.

Toyota might have great reasons for opening a chain of premium quality gas stations, but the second they required me to use them, I'd never buy another Toyota for as long as I lived.

I want to bring my own drives, just as I have since I bought my first DS-412+ 13 years ago.

[numpad0]

People are going to ignore that and leave bad reviews online, which will have compounding effects. SMR drives works in RAID until CMR buffer regions are depleted, and then RAID starts falling apart. This will undoubtedly create wrong impressions that Synology products, not the drives, are not trustworthy.

[Melatonic]

What if the device had a minimum benchmark feature that would test any new drive ? And fail the worst ones ?

[numpad0]

SMR drives work like SSD: writes are buffered to CMR zone, consolidated into an SMR track data, copied into onboard cache RAM and written to SMR zone. SMR tracks has sizes of 128MB or so, and can be written or erased in track-at-once manners by the head half-overwriting data like moving a broad whiteboard marker slowly outward on a pottery wheel, rather than giving each rings of data enough separation. This works because the heads has higher resolution in radial direction in reads than writes; the marker tip is broader than what the disk's eyes can see.

This copy operation is done either while the disk is idling, or forced by stop responding to read and write operations if CMR buffer zone is depleted and data has to be moved off. RAID softwares cannot handle the latter scenarios, and consider the disk faulty.

You can probably corner a disk into this depleted state to expose a drive being SMR based, but I don't know if that works reliably or if it's the right solution. This is roughly all I know on technical side of this problem anyway.