HGST / WD Firmware Updates

zbyte

Dabbler
Joined
Jan 3, 2021
Messages
35
I searched the forums but could not find a specific thread for this specific question.

Building a new TrueNAS system, I have mixed drives, Seagate, WD and HGST.

Seagate's when tested with "smartctl -t long" reported errors, following some threads here and others I found out there were issues sometimes with the drive overshooting on its seek requests and that was reported (by Seagate algorithm) as errors. They (Seagate) use their own different metrics and reporting. Anyway, I was able to download and apply the specific firmware to fix this error reporting, and I think even will reset the counters, haven't tested yet.

But when I tried to download the latest firmware for WD and HGST drives I could not find any and they direct you to open a support case, they will take more than 1 week to answer and they only do it when you "bump" the support case.

Listing below my case, maybe anyone can help or maybe I should go ahead with whichever firmware is included.

Initial question
Setting up these drives on a NAS and would like the latest working approved firmware for these to guarantee I won't find any inconsistencies please.
MODEL - SERIAL
1 - WD40EZRX WCC4E0HXET2Y
2 - WD40EZRX WCC4E5UEVHAU
3 - WD40EZRX WCC4E1ERAVVE
4 - HDN724040ALE640 PCGXNYLS
5 - HDN724040ALE640 PCG7SDYB
6 - HDN726060ALE610 NAG18MKP
7 - HDN726060ALE610 NAG17Z2P

Note I'm listing the actual models and serials from the drives

Their answer

Dear OMITTED

Thank you for contacting Western Digital Technical Support. My name is OMITTED

It is my pleasure assisting your case and here is the case number for future references: 210208-000471

I understand your concern for your Western Digital drive firmware version issue. Based on your concern I request you to please share the images of the drive label with model number and serial number visible on it. As we checked the serial number you have mentioned it is showing up as incorrect

If you have any further questions, please reply to this email and we will be happy to assist you further.

Sincerely,
omitted
Western Digital Technical Support
http://support-en.wd.com

So I had to upload all images of the actual drives, unmount them and mount these back but I did it

I followed up several times, finally today their answer

Dear omitted,

Thank you for your continued response. I do apologize for any trouble you may be having. Based on the Serial Numbers you provided all of these drives are either our WD Green or HGST DeskStar drives. All of these drives are released with their latest firmware release, and we do not have any firmware updates for these drives.

If you have any further questions, please reply to this email and we will be happy to assist you further.

Sincerely,
omitted
Western Digital Technical Support
http://support-en.wd.com

I replied with this and escalated, not that I think they will take me seriously

Hi omitted, thank you for your answer, it really doesn't provide a satisfactory response, I have been in the technology business for many years, when the drives were 5 mbytes in size and manufactured by Winchester, WD was top of the line and I used those continuously. Anyway modern drives are released by all manufacturers and on the field they found problems with the drives sometimes overshooting on seeks or with specific bugs on the firmware release. If you tell me the Greens which is a consumer drive fine but the HGST's which are not consumer drives at least not marketed like that and sometimes used on OEM equipment like Dell, HP or what have you I know for a fact that these drives are usually reviewed and new firmware provided for them, many times through the OEM. Now if it is WD policy not to provide the firmware directly to customers I could understand that but I will consider my drive choices (and others too) more carefully in the future. Seagate does it and you don't have to contact support to download the latest firmware including the changes and why it is recommended to update.

In your experience, really they only develop one single firmware? no revisions? no problems in the field? (which I'm sure they would have)

Just curious, because I really wasn't expecting these answers from WD

Thank you all
 

ChrisRJ

Wizard
Joined
Oct 23, 2020
Messages
1,904
IIRC the WD Greens are usually not recommended for NAS use. With this number of drives in particular I would start getting worried about vibrations.

Is there a specific reason why you want to update the firmware on those drives?
 

zbyte

Dabbler
Joined
Jan 3, 2021
Messages
35
There is this thread
Since these are low RPM's looks like are ideal for home NAS, low energy and such, as long as you change the parking idle timing you should be good (or so it seems)

I just found out about the Seagate issues with firmware and since I know not only Seagate but all drives manufacturers for mechanical drives and SSD usually discover problems in the field corrected with new releases I thought on asking WD, obviously their policies seems to be different and I came here to the forum to find out based on previous experiences.
 

joeschmuck

Old Man
Moderator
Joined
May 28, 2011
Messages
10,970
There is this thread
That thread came about because drives were parking the heads too frequently and the warranty for head loading count was fairly low (WD has increased the head load limits since then). I set my REDs to never park and they lasted 5 years before I swapped them out with larger drives. This was not a firmware change for WD, it was a TOOL if you will to change a few values within the drive from the default values. I have seen new firmware for some drives and SSDs to enhance or correct a deficiency.
Seagate's when tested with "smartctl -t long" reported errors
Exactly what errors? Are you speaking of the ID1 Read Error Rate? With Seagate that data typically needs to be manipulated, only the first 4 bytes relate to read errors. Use the command smartctl -a -v 1,raw48:54 /dev/ada0 for a designated drive. Reference this thread for more information.
 

zbyte

Dabbler
Joined
Jan 3, 2021
Messages
35
That thread came about because drives were parking the heads too frequently and the warranty for head loading count was fairly low (WD has increased the head load limits since then). I set my REDs to never park and they lasted 5 years before I swapped them out with larger drives. This was not a firmware change for WD, it was a TOOL
Thank you, I'm aware it is a TOOL but there is some extensive reference to the features of such Green Drives by cyberjock and used with TrueNAS on that thread too and apparently they are onpar with the RED's (which are recommended for NAS use), actually these won't suffer from SMR whether the RED's do (not the Plus version)
Exactly what errors?
The errors you referenced on another thread :), not the specific Ironwolf thread you linked, another thread (which I can't find now), you basically stated the ID's to look for and the ID's to ignore, there is even a reference (out of these forums to another site) on someone describing and analyzing how Seagate gets to these errors, drive heads overshooting their target and such, I found the firmware that fixes these conditions for Seagate (so no errors are reported), and figured WD would have "newer" firmware (not specifically for the same conditions as Seagate but others)

I will post the thread and specifics if I find it, it was quite useful
 

joeschmuck

Old Man
Moderator
Joined
May 28, 2011
Messages
10,970
actually these won't suffer from SMR whether the RED's do (not the Plus version)
I don't understand your point with respect to firmware upgrades. Also the REDs were CMR originally, WD started making SMR Reds but that was years later.

The Hard Drive Troubleshooting Guide that I created I tell people to ignore the Read Errors (ID1) because those are not errors that mean your drive is failing. At the time of the writing I did not know how to decode the Seagate values either. But i will stick to my guns, the value does not matter when troubleshooting hard drive failures. The Ironwolf is a Seagate drive as well so the command should work fine.

and figured WD would have "newer" firmware
That is an assumption and you know what they say about that... Most companies will not put out a fix unless there is some adverse reaction happening, mainly one that hurts their bottom line $$$$$$.
 

zbyte

Dabbler
Joined
Jan 3, 2021
Messages
35
@joeschmuck I'm just replying, not intending for this to be an argumentative discussion

I don't understand your point with respect to firmware upgrades
My point is the same you reflected with your comment here
I have seen new firmware for some drives and SSDs to enhance or correct a deficiency.
The same way it corrects deficiencies on SSDs it does for mechanical drives.

Also the REDs were CMR originally, WD started making SMR Reds but that was years later
My answer was more directed to a previous comment by @ChrisRJ mentioning the greens are usually not recommended which might be accurate from the "recommendation" standpoint but cyberjock (empirically) used plenty of these with very good results based on his thread. If I have the drives and these test good and TNAS will allow me to run a resilient pool why would I discard these? They (the Greens) were all CMR whether the same size Reds (even the 2TB Reds) were SMR around the same date 2014/2015, now if I'm incorrect on these dates I'm sure you could provide some correction.

The Hard Drive Troubleshooting Guide that I created I tell people to ignore the Read Errors (ID1) because those are not errors that mean your drive is failing. At the time of the writing I did not know how to decode the Seagate values either.
That is correct, thank you for the guide, it is very helpful, I read it before, and extracted these (your exact words):

"If ID’s 5, 197, or 198 have any value greater than zero (0) then there has been some defect identified in the media."

Any other counters in the case of Seagate's specifically could have different than zero values but not been really "an error" like for example ID 1 Raw_Read_Error_Rate, which every other manufacturer should be zero but for Seagate is common to accumulate.

I went further and "assumed" a firmware update will correct the issue if there was any, and I found something related for my Seagate drives.
firmware version GA10 for Seagate
It doesn't specifically defines what an "unexpected SMART trip" is though, there I found several releases (at least 4) of firmware's for the same drives. Also several firmware releases show the same fix so I guess they did not get it right initially or something else was fixed.

So I (again) assume other manufacturers probably were actively correcting mishaps on software, bugs and such that could affect their firmware's. What really made me to post here was the pragmatism from WD that they don't update their drives, and you comment:
Most companies will not put out a fix unless there is some adverse reaction happening, mainly one that hurts their bottom line $$$$$$
Is right on the $$$$$$

But noted that for example Dell (an OEM) will support and provide firmware for their drives even if these are manufactured by Seagate, WD, HGST or else, sometimes the drives they provide are without relabeling, sometimes they relabel the drive but leave it as it is, sometimes not only relabel but they request a geometry change from the manufacturer like different number of cylinders and such (years ago I remember a Sun Microsystems drives like this)

I got my pool up and running with updated Seagate firmware drives, HGST and Greens, currently running smartctl -t long on all drives

Now let me ask you @joeschmuck I set the parking to 300 seconds (for the Greens), do you think not to park would be better since this is what you are using?

Thank you for your feedback
 

jgreco

Resident Grinch
Joined
May 29, 2011
Messages
18,681
The days of geometry changes are in the far past; these days, disks no longer have spindle sync or CHS addressing, but rather use LBA. As a result, I suspect the number of firmware fixes required has dropped significantly, and with drive manufacturers seeing the writing on the wall, basically all of them have been bought out by Seagate/WD/Toshiba, who have been dropping product lines, meaning that there's a lot less to get wrong.
 

joeschmuck

Old Man
Moderator
Joined
May 28, 2011
Messages
10,970
Now let me ask you @joeschmuck I set the parking to 300 seconds (for the Greens), do you think not to park would be better since this is what you are using?
I think that depends on your usage. Let me explain how I would determine it. If the drive is accessed frequently (less than once every 5 minutes) then the drive heads would never park and that is not an issue. If the drive is accessed maybe once every 6 minutes then the heads are going to unload and then after 1 minute load again. This would be not good. And if the drive were to be accessed once every 10 minutes, the same result is the heads would be loaded very often. But if your hard drive is accessed once every 2 hours then I would say a 5 minute timer is okay, in my opinion. Use your SMART data to look at the head load count, watch the count go up over a 24 hour period of time. If you are comfortable with how much it increments then you are fine, but if you feel it's incrementing too much then I'd disable the head parking and let the drive run. Just so you know, when you shutdown the system the drive will automatically park the heads.
disks no longer have spindle sync or CHS addressing, but rather use LBA.
Ouch, my brain remembers those old days, fun days when learning about how these things worked.
 

zbyte

Dabbler
Joined
Jan 3, 2021
Messages
35
But if your hard drive is accessed once every 2 hours then I would say a 5 minute timer is okay
Thank you, very helpful.
Ouch, my brain remembers those old days, fun days when learning about how these things worked.
Yeah I remember entering Cylinder and Heads inside the BIOS? I don't even remeber the BIOS auto detected the drives as today. My first hard drive was a Winchester External 5 MB drive, was it a Centronics connector? considering my floppy drives were 360 KB, 256 KB RAM and an MDA adapter. 1984, that 5 MB drive was like WOW, 2nd drive was a Maxtor? MFM drive I think .... good times. Then 640 KB was standard on the IBMs, extended 384 and QEMM (Quarterdeck Memory Manager), I bet you guys remember that.
 

ChrisRJ

Wizard
Joined
Oct 23, 2020
Messages
1,904
It sometimes feels like all the "old-timers" have gathered here :wink:. One of the reasons I like this forum a lot ...
 

joeschmuck

Old Man
Moderator
Joined
May 28, 2011
Messages
10,970
Then 640 KB was standard on the IBMs, extended 384 and QEMM (Quarterdeck Memory Manager), I bet you guys remember that.
I most certainly do remember that and ran that. QEMM, my memory might be failing me but I think I had a RAM card with 4MB. It was a full length card and fully populated with RAM chips. I want to say it was about 26 years ago, I know my daughter wasn't born yet and she's 24. God I'm old!
 

Constantin

Vampire Pig
Joined
May 19, 2017
Messages
1,828
Last edited:

ChrisRJ

Wizard
Joined
Oct 23, 2020
Messages
1,904
What computer was that?
 

jgreco

Resident Grinch
Joined
May 29, 2011
Messages
18,681
Hah. but could your hard drive sing like my floppy drive?

My first computer coded only in ASM and used toggle switches + a alphanumeric LED display. Keyboards? Pure luxury!

Alphanumeric LED display? Seriously? Real men used rows of LED's. :tongue:

Actually I don't really understand the Altair fans out there. Was it really so hard to bap some encoder/decoder logic in there to drive a keypad?

I still have a Synertek SYM-1 here. :smile:
 

Constantin

Vampire Pig
Joined
May 19, 2017
Messages
1,828
What computer was that?
So i could not find it offhand but the Profi-50 reminds me of it though I seem to remember toggle switches on mine and the case was white plastic not black metal.

I presume it was a very obscure German-made machine that my dad thought was a better educational tool than the C128 I acquired a year or two later.:smile:
 
Last edited:
Joined
Jul 2, 2019
Messages
648
Lordy... I am getting old. I can remember the Altair 8800 in, I think, Byte magazine - when it was on paper and you had to buy it at the newspaper store. Where my dad worked, they started automation with Digital (DEC) machines so he asked one of the technicians what type of computer he should buy for his son (me). Of course, a DEC Rainbow was what was suggested and that is what he bought. But there wasn't much software (or maybe, more correctly, software that was useful to me...). Shortly thereafter he bought me an Apple //c because it was more "educational". I taught myself assembler using a compiler called Merlin. Of course, Aztec and F-15 Strike Eagle were much more fun - even on that green monochrome 9" Monitor //c. And the Apple Scribe thermal printer - must have been the most expensive ribbons ever.

I used to have the maintenance "briefcase" for the old DEC VAX 8800 - with the white lab coat, torque wrench and insulated hook pole in case someone fell into the chassis - 387V if I remember correctly. I miss the DEC Alpha - too bad Compaq killed off the Alpha...
 

Constantin

Vampire Pig
Joined
May 19, 2017
Messages
1,828
Yeah, I'm wondering if my first machine was some sort of local adoption / adaptation of the BBC Acorn. It's long gone.

I also jettisoned my C128 and Amiga when I left for the states and the oldest CPU I now own is a Mac SE. It should be in running condition though it likely needs a new battery for the clock.
 
Last edited:

ChrisRJ

Wizard
Joined
Oct 23, 2020
Messages
1,904
Commodore 128 was also my first computer. Apart from gaming I did a lot of programming with BASIC and also Turbo Pascal 3.0 on CP/M (although with 2 MHz the Z80 was really slow).
 

joeschmuck

Old Man
Moderator
Joined
May 28, 2011
Messages
10,970
OMG, Altair brings back memories. We had one that had a voice synthesizer in it, it was fun programming it to talk but you needed to do it phonetically.

My first computer coded only in ASM and used toggle switches + a alphanumeric LED display. QWERTY Keyboards? Pure luxury!
I can relate to that, the first computer I ever got to play with I had to build from a kit and it was two hex LED displays and a Hex Keypad. It had a few discrete LEDs you out make light up as well. I don't recall anything more about it. That was in 1976. After that it was an IBM Mainframe in 1977 (punched cards) and a TRS-80 with 16KB or RAM.
I miss the DEC Alpha
I was working on a DEC Alpha 3000 about I think 7 years ago (seems like yesterday), it was very active and fortunately we retired it because finding the motherboard for it as a repair part was terribly difficult. It was the only part that seemed to fail during it's last few years of service.

Funny how a thread on hard drives becomes a thread about our fond memories.
 
Top