Western Digital Device Analytics - 'Power On Hours' warnings on WD Reds after just three years of NAS use

[andrewilley]

Had a bit of a panic yesterday when my DiskStation sent notifications that two of its 6TB WD Red NAS drives (SHR-array) had gone into Warning state at the same time, advising doing S.M.A.R.T. testing on both drives.

Digging a bit deeper, it seems that WDDA (Western Digital Device Analytics) has flagged a power-on-hours warning, and is suggesting replacing both drives. As they were installed at the same time, they both had about the same run times - approx 26,400 hours. Strangely enough, that's just a day or two after WD's Three-Year Warranty ran out - so WD appear to be suggesting replacing their devices as soon as they drop out of warranty.

Extended S.M.A.R.T. testing shows no errors at all on either drive so I'm not planning on replacing them both just now, but how do I disable just the run-time warning in WDDA, leaving all other WDDA testing features active? Currently they are both still showing as 'Warning' status.

Andre

 

[Coop777]

If your NAS is still under warranty, this would be a good question for Synology Tech support to give you any guidance.

 

[EAZ1964]

I can recommend to switch off the nas in the night, easy 30% power use reduction and run hours reduction.

Another thing is, the old rule to spread your risk, by mixing disks with different age and/or brand in the raid to lower the risk for simultaneous failure.

 

[andrewilley]

Backups and some file transfers run overnight, so network and NAS stays powered up 24/7. NAS are generally designed to run continuously anyway.

Not really expecting any issues with the drives after three years, and if so hopefully they'll die one at a time (the other two drives in the NAS are much newer to spread the danger). Was really just trying to see how to disable the 'Warning' status for just operating hours, while keeping other more specific and critical warnings available.

Andre

 

[EAZ1964]

Understand your use case.
Although NAS can run 24/7, they certainly do not need to and it saves money and resources to switch them off if possible (typically 8hrs a day switch off saves 60$/year with my electricity cost, easy money) .

 

[Tedster]

I swapped out my 3tb WD Reds after about 5 years service. One of the original 5 drives failed, just outside of warranty.

The rest are still in use in various PC's around the place, as storage drives.

-- post merged: 20. Mar 2023 --

Oh, some of them are approaching 60,000 hrs now.

 

[Coop777]

I haven't seen as many threads on it lately, but over the years on this form and the community there have been quite a few WD Red drive failures posted.

 

[durin]

I haven't seen as many threads on it lately, but over the years on this form and the community there have been quite a few WD Red drive failures posted.

I'm sure anecdotal observations of drive failures of all the main HDD manufacturers can be found online. Is there evidence that WD Red drives are failing at a disproportionately high rate? Genuinely curious as I'm considering some Red Plus drives as I gradually increase the capacity of my NAS, not least due to the notably lower power consumption of 5400 rpm drives.

 

[fredbert]

My limited experience (due to not so many issues) has been good, so far. Below the WD Red are the older ‘Red’ which are CMR technology, akin to today’s Red Plus. Approximate hours.

DS215j: I initially got two WD Red 3TB. After 31,000 hours one was reported as failing. I swapped it and now have this updated pair running for a further 36,000 hours (currently as two Basic drives).

Later I did an extended SMART test on the failing drive, after out-of-NAS tests and zero wipes, and it came back as healthy. It’s a temp disk now.

DS218+: two WD Red 4TB no issues so far, around 35,000 hours.

DS1520+: five Ironwolf 8TB, 4 in SHR plus a basic. No issues so far.

All NAS protected using UPS.

I’ve had more drive failures will ‘desktop’ variants. Over the years: Hitachi, Seagate, and WD. The latest being a 6TB WD Blue from an Elements enclosure. So I have a stack of magnets culled from giving platters a comprehensive reformatting using the ice-skating method on the garage floor and then a vice.

 

[andrewilley]

Synology Tech Support have just replied to my Ticket saying it's not their problem, and they can't/won't filter the results that WDDA provides:

Unfortunately this appears to be an issue that we would not be able to assist with.

That information is pulled directly from the drive and as such, you would need to contact Western Digital themselves for assistance with how to stop that reporting if it is even possible (I would suspect they wouldn't want that to occur).

On our end, that would require modification of system files and that is not something that we would support nor assist with I am afraid.​

So now to try WD Tech Support...

Andre

 

[Deleted member 5784]

WD Tech support are likely to just tell you - 'Everything is working as designed'...

This apears to be another instance where SMART reporting is being used to upsell HDDs by an unscupulous manufacturer. As documented here & elsewhere, Synology have manipulated the SMART system to issue FUD drive warnings; it seems that WD are now doing the same.

Coming at this from a different angle, what is the symptom on the NAS from all this - how and where is the warning being displayed? Is it simply ignorable? What smartID is this showing up under?

 

[SynoMan]

“Clearly predatory”: Western Digital sparks panic, anger for age-shaming HDDs

Drives automatically get a "warning" flag if powered on for 3 years.

arstechnica.com

 

[Telos]

TLDR:
sudo synosetkeyvalue /etc.defaults/synoinfo.conf support_wdda no

 

[andrewilley]

I agree it's easy enough to simply turn off WDDA monitoring (can be done via the Synology control panel too) but that then means I will not receive any genuine (i.e. not simply sales/marketing pitches) warning messages which may occur in the future, making the system completely pointless.

Andre

 

[fredbert]

I inferred that WD and DSM still combined to provide SMART, even with WDDA switched off. So not completely in the dark.

 

[Telos]

A major concern involves pool repairs. If a drive fails and there is a warning attached to the remaining drives, my understanding from that article, is that a repair/rebuild is prevented by Storage Manager.

It's interesting too to read that Synology has walked away from WDDA in newer models, but is leaving WDDA intact elsewhere.

I will not receive any genuine (i.e. not simply sales/marketing pitches) warning messages which may occur

I don't believe that is correct. Synology SMART will still be in play. WDDA is akin to the Ironwolf health mess that Synology rolled out years ago, only to later abandon it within 2 years. Deja vu!

 

[andrewilley]

I inferred that WD and DSM still combined to provide SMART, even with WDDA switched off. So not completely in the dark.

Yes, S.M.A.R.T. will (and does) still work even with WDDA turned off. But I have to assume that WDDA is meant to do more than just scare people into buying new drives every three years.

Andre

 

[Telos]

But I have to assume

Either there are specific advantages or not. If one can't identify the "advantages", then one is soley acting out of FUD/FOMO.

As Synology hasn't included WDDA in newer units, apparently they consider it non-essential.

 

[Deleted member 5784]

Yes, S.M.A.R.T. will (and does) still work even with WDDA turned off. But I have to assume that WDDA is meant to do more than just scare people into buying new drives every three years.

Andre

It's a gimmick, devised for the purposes of 'product differentiation' or somesuch BS; the rest of the world - including the world of enterprise - relies on SMART data alone and gets by fine without WDDA.
And if the gimmick is actually causing false alerts & panic, then it's doing more harm than good. Switch it off and rely on SMART.

 

[Coop777]

This is yet another shady business practice by WD. Completely unacceptable to basically highjack the hardware. Key thing now is to switch it off in DSM before the drives get flagged.