Just saw this
Curious if other models are impacted. Has anyone exceeded 65k hours on 5100, 5200 or 5300?
Curious if other models are impacted. Has anyone exceeded 65k hours on 5100, 5200 or 5300?
Micron 5200 firmware bug
- Thread starter ca3y6
- Start date
Notice: Page may contain affiliate links for which we may earn a small commission through services like Amazon Affiliates or Skimlinks.
Crikey; I've got at least some 5200s that'll have been running pretty much non-stop for 6 years or more...
<clickety click>
Oldest has a power-on count of 6.29 years/55,000-odd hours, all are on the D1MUx20 firmware version (never bothered to upgrade to x30, if it ain't broke...).
Given that the replacement 5400 parts are currently selling for stupid money and are unavailable even with a third mortgage and a kidney, surely they're in for a world of shit if they won't release a FW update for this, assuming it's real...?
Exactly how many SSD firmware bugs have we had now that are down to integer overflows? Surely this has had to have been a part of QA testing for over a decade now...? Was there ever such a huge problem with (presumably) SMART counters in the ol' platter-based HDD days?
Edit: just found one of evidently my first 5200's with a power-on count of 62269 hours... so, 136 days remaining. Gulp. Bought in october 2018 according to my records.
<clickety click>
Oldest has a power-on count of 6.29 years/55,000-odd hours, all are on the D1MUx20 firmware version (never bothered to upgrade to x30, if it ain't broke...).
Given that the replacement 5400 parts are currently selling for stupid money and are unavailable even with a third mortgage and a kidney, surely they're in for a world of shit if they won't release a FW update for this, assuming it's real...?
Exactly how many SSD firmware bugs have we had now that are down to integer overflows? Surely this has had to have been a part of QA testing for over a decade now...? Was there ever such a huge problem with (presumably) SMART counters in the ol' platter-based HDD days?
Edit: just found one of evidently my first 5200's with a power-on count of 62269 hours... so, 136 days remaining. Gulp. Bought in october 2018 according to my records.
Last edited:
Presumably there's at least a handful of OEM versions of these drives out in the wild as well...? Is anyone aware of any other firmware releases for them maybe?
I can see Dell at least published a FW version for the 5200s back in 2024 but presumably that's for one of their custom versions and not for retail (and in any case its changelog doesn't match up with the potential 65k bug).
I can see Dell at least published a FW version for the 5200s back in 2024 but presumably that's for one of their custom versions and not for retail (and in any case its changelog doesn't match up with the potential 65k bug).
I have two and they are MP32. I remember there being an update (probably MP33) when I first got these drives, but I guess I never updated/managed to get it to work. I think it might've been because I saw that it required a lenovo server.
The micron branded 5300 max I have, I was able to update to the most recent firmware using the micron storage software a while ago. I remember not being able to get it to work in proxmox and didnt want to pull the 5300 pros to try in a different machine. Not sure if that one is able to update the firmware for a lenovo branded drive though.
All of mine are 20k - 40k POH so I guess I'm safe for a while while yall sort this out.
The 5300 seems to still be supported and a firmware is available for download. The guy on reddit says the 5100 isn't affected so chances are the 5300 may not either (though with Samsung the whole SAS product line was affected). So will monitor the micron website for a new 5300 firmware version showing up before worrying about lenovo.
Actually I think I found a way to download the lenovo firmwares. Just lookup your drive in parts lookup, look for compatible systems, take a random one, then go into the datacenter support, search that system, then under Lenovo Storage Linux Firmware, you can download a big "bin" file that you can unzip, (using 7zip) and a subfolder contains all the firmwares. Inside the utility I think is "fdrvwl flash" but you can probably also use OneCli to flash them (haven't tried). If you edit "flashdrv.bin" in notepad, you can see the mapping of drive model to the firmware version (search your model), for the 5300 MP33 seems to be the latest version.
Last edited:
Well there was another post on the reddit thread today saying;
Has anyone tried giving it a whirl yet?
If I'm reading that right - and I'm not sure I am - it sounds like they were having the issue on their 5200s already, but managed to fix it by upgrading the firmware to the D1MUx30 i.e. the last one published (circa Jan 2021 based on the dates within the zip).Hi, funny thing... on our Supermicro Server with 6 - Micron 5200 ECO 480GB "MTFDDAK480TDC Drives" we had the same strange behavior... Speed is 1 MB/s... Micron Tool sayed: Everything Fine... could you please share your Firmware Version?
Upgrade today was from D1MU004 -> D1MU030...
Thanks for your answer
Has anyone tried giving it a whirl yet?
You can go to Micron's website and register using a non-Yahoo, -Gmail, etc. address and download the latest FW for their drives. The 5200's last release was in 2021.
Go to www.micron.com/search-results?activeTab=search-tech-docs then enter "5200" or whatever then sort by date. Click on the Get Access link and you can register.
I purchased some 5210s and they can also be downloaded. The mapping is included in the combined ZIP file(s) under firmware.properties.
Go to www.micron.com/search-results?activeTab=search-tech-docs then enter "5200" or whatever then sort by date. Click on the Get Access link and you can register.
| Firmware | Drive Model; Capacity |
| D1MU030 | 5200ECO, 5200PRO, 5200MAX; 240GB, 480GB, 960GB, 1920GB |
| D1MU430 | 5200ECO; 3840GB |
| D1MU530 | 5200PRO; 3840GB |
| D1MU830 | 5200ECO; 7680GB |
I purchased some 5210s and they can also be downloaded. The mapping is included in the combined ZIP file(s) under firmware.properties.
This happened across multiple (ssd) manufacturers using different nand controllers, I'm wondering what they have in common and if they use a shared "reference" design + firmware that uses 16bit for data?
Micron 5100/5200/5300 is Marvell 88SS1074 or similar. Highly doubt the big three Dell HPE Lenovo made any in-house design. As per usual just different sticker, different part number, and different firmware.
Last known to me firmware for original Microns:
5100
5200
5300
Last known to me firmware for original Microns:
5100
[D0MU075]
drive_model=5100ECO,5100PRO,5100MAX
firmware_type=MU
capacity=240,480,960,1920
display_version=D0MU075
[D0MU445]
drive_model=5100ECO,5100PRO
firmware_type=MU
capacity=3840
display_version=D0MU445
[D0MU845]
drive_model=5100ECO
firmware_type=MU
capacity=7680
display_version=D0MU845
5200
[D1MU030]
drive_model=5200ECO,5200PRO,5200MAX
firmware_type=MU
capacity=240,480,960,1920
display_version=D1MU030
[D1MU430]
drive_model=5200ECO
firmware_type=MU
capacity=3840
display_version=D1MU430
[D1MU530]
drive_model=5200PRO
firmware_type=MU
capacity=3840
display_version=D1MU530
[D1MU830]
drive_model=5200ECO
firmware_type=MU
capacity=7680
display_version=D1MU830
5300
[D3MU001]
drive_model=5300BOOT,5300PRO,5300MAX
firmware_type=MU
capacity=240,480,960,1920
display_version=D3MU001
[D3MU401]
drive_model=5300PRO,5300MAX
firmware_type=MU
capacity=3840
display_version=D3MU401
[D3MU801]
drive_model=5300PRO
firmware_type=MU
capacity=7680
display_version=D3MU801
Well, I've bumped two of my 5200s to D1MUx30 (again, didn't have any difficulty obtaining it at the time and as ca3y6 says it looks to still be there behind the loginwall) to no ill effects; I've usually always done FW upgrades "offline" before but the upgrade via msecli appears to work completely transparently when the drives are up and running.
But I think the unanswered question is does D1MUx30 actually fix the potential problem here...? Probably we're going to need to wait for GermanPCBHacker to let the thread know what "fix" version they were given by Micron, or for a StHer running D1MUx30 past 65k hours to confirm the issue is/isn't fixed/real...
But I think the unanswered question is does D1MUx30 actually fix the potential problem here...? Probably we're going to need to wait for GermanPCBHacker to let the thread know what "fix" version they were given by Micron, or for a StHer running D1MUx30 past 65k hours to confirm the issue is/isn't fixed/real...
Indeed, it's been over a highly varied range of kit IIRC so I do wonder if instead of being down to a shared factor somewhere, it's instead down to the universal constant of software bugs/lack of QA
OP of reddit thread replied back that they were given D1MU040 for their Micron 5200s, so the bug is most likely real if they were given firmware newer than what is publicly available.
Last Dell firmware for the 5200 series (F005) was compiled back in 2021, so I guess the Dell version and all OEM subversions are probably affected as well.
Last Dell firmware for the 5200 series (F005) was compiled back in 2021, so I guess the Dell version and all OEM subversions are probably affected as well.
Well it's a real pisser if there really is a D1MU040 knocking about that they're not releasing... presumably there's also a corresponding set of D1MU440/D1MU540/D1MU840 as well to cater for the other drive sizes.
Will have to see if the original poster decides to post the file(s) or not I guess.
Will have to see if the original poster decides to post the file(s) or not I guess.
How about calling the CEO of Micron?
Edit: I emailed Micron's CEO via the corporate secretary, requesting publication of the firmware on the web site. Will he read it, do it, no idea.
Edit: I emailed Micron's CEO via the corporate secretary, requesting publication of the firmware on the web site. Will he read it, do it, no idea.
Last edited:
Another user on the thread, EtHeO18, has posted hashes of at least D1MU040 and D1MU840 from what I guess is the whole 5200_D1MU_040_440_540_840_fwbin.zip bundle so there's at least hope that someone out there has the whole nine yards and will pluck up the courage to post if micron seem unwilling.
Though Micron so far has a pretty good track record at publishing their firmware updates. If they just discovered the problem, and issued an emergency patch, it is possible that they are going through some QA before releasing the updated firmware on their website and to the other OEM.
I have submitted a new ticket with Micron support, hopefully they say something.
if they refuse to supply the firmware, i will be letting them know i will contact the Consumer Protection Financial bureau, Federal Trade Commission, and Better Business bureau about not correcting a known defect when the correction is known to be available and works.
if they refuse to supply the firmware, i will be letting them know i will contact the Consumer Protection Financial bureau, Federal Trade Commission, and Better Business bureau about not correcting a known defect when the correction is known to be available and works.
Code:
There is a known firmware issue in firmware versions D1MU004, D1MU020, and D1MU030 that is corrected in a NON-PUBLIC released version of firmware D1MU040 for the 5200 series drives. The issue in older firmware's occurs when the SMART power on hours parameter exceeds 65535 hours.
At the point the counter exceeds 65535 hours, the drive becomes near unresponsive with a data throughout of 1 MB/S or less.
It has been confirmed my multiple people that this issue occurs and it has been confirmed that version D1MU040 corrects the issue. It has been confirmed that the following md5sum hash of the firmware .bin files are available
e8e51fcac5d602685233a4c53b355bf3 crucial/5200/D1MU040/1.bin
d7c3cc2a711635887c7a8c181ad5e8f4 crucial/5200/D1MU840/1.bin
I am requesting a copy of this new D1MU040 firmware to allow me to upgrade my multiple ECO 5200 1.92TB drives as they are NEAR the 65535 hour mark and we cannot allow the system to crash due to the firmware issue.