[HELP] Internal HDDs Stop Working (Cannot Write) Some Time After Login

[greenfrog5]

This has been an intermittent problem for me since I updated from Chimera+Yosemite to Clover+Catalina. This update has caused all sorts of problems (learning Clover, unsupported hardware, etc) for me, which I have now mostly worked out. I still get sometimes freezes running Catalina on my SSD, but have been able to run for multiple hours, and several days much of the time. Despite fixing my Clover and other stuff, I continue to experience this HDD issue:

I have a number of partitions on internal HDDs (generally HFS+), and most have experienced this at times. It tends to be certain ones more often, or first. What happens: I boot and login, and HDDs work normally. Over time (usually an hour, or several), the internal partitions stop working (often one-by-one). I lose the ability to write/edit to them with an error "This operation cannot be completed. Error -50". I can still read from them, but they are essentially read-only. Often it is one, or some (but not all) of the partitions on a particular physical drive. I feel that if I run long enough, they all might go read-only over time (when this is happening). When I was booting off of one of the partitions, that partition (APFS, Catalina) wouldn't be affected (i.e. the system would function/run ok), but the others on the same physical drive would go read-only (error -50).

The only recovery is to reboot. When booting Catalina off my non-SSD recently (before fixing Clover, boot from SSD), there were times when this did not happen (many days uptime without issue), but it would occasionally occur at times. The issue seems not to happen for many hours or days (though my uptime has been limited lately, so I don't have a lot of long-term tests with my current better-working system). It has been reoccurring more frequently now running off my SSD with new Clover setup (which otherwise runs pretty well, except from some persistent freezes still).

I cannot find anything online about error -50, or similar sounding problems. I don't know what to try, or where to begin, so please let me know what I can look into on this one.

 

[greenfrog5]

I'd like to re-ignite this thread in hopes of some assistance. I have been troubleshooting a variety of issues with this machine, and am getting closer to stable performance. However, I have not made much forward progress with this hard disk issue. I have, however, gathered more information which hopefully can help someone help me sort this out.

Two of my physical disks are affected by this issue. All of the partitions on these two disks are generally affected. Some happen consistently within about 30 mins. Others as long as an hour or more. It happens on every boot now (more than before).

A new HDD I recently purchased and formatted is not affected by this issue.

I have tried changing SATA cables, SATA ports etc. I finally emptied one of the problem drives off and reformatted it. The newly formatted drive experiences the same issue. Placing this drive in an external enclosure- it does NOT experience the issue. I have tried erasing caches and clearing NVRAM (per online recommendations), which seemed to help at times, but not always, and has not conclusively resolved the issue.

Shortly after boot when the partitions are working correctly, Disk Utility First Aid runs and passes without issues. Once the problem occurs, Disk Utility cannot unmount the drives, and cannot run First Aid.

Is this a problem with the partition map? I was very unpleasantly surprised that formatting the disk didn't fix the issue. Also, the disk working in an external enclosure suggests it is not a physical/hardware problem with the disk itself. It is a 4TB drive, and I have not been able to figure out how to change the partition scheme from GUID, in hopes of erasing the issue, then changing it back.

 

[Edhawk]

Doesn't the way the new drive works without issue, tell you that your existing drives are failing?

 

[greenfrog5]

The problem drives are similar (identical?) HGST 4TB drives. I was also able to copy data for several hours from the internal drive to the USB dock drive. Keeping the internal drive busy prevented the error -50/read-only from occurring for that time. Within 30 minutes of inactivity, the internal drive went -50/read-only. The same partitions tend to go first (not all at the same time), but given enough time, they all go.

It is as if the partition half-goes to sleep and can't be woken back up for data writing. There are a few online reports of drive firmware (typically external, and WD brand) causing spindown that OSX cannot fully revive leading to error -50. Interestingly (coincidentally--or maybe suspiciously), one recommended solution for this spindown issue is to disable "put hard drives to sleep" which is what I just had to do to fix my 30-min reboot issue on my SSDs. Already having it disabled, but am still getting -50/read-only issue.

The read/write/data on the drive seems to "work" except for when this issue occurs. These are the same drives I was using with Yosemite and I'm pretty sure this started happening right when trying to update to Catalina.

I don't dispute there is something "wrong" with these drives, but I'm not convinced it is simply drive failure. They seem more incompatible, though there aren't widespread reports online or anything. I still have not tried writing secure data to the entire drive, which some suggest.

Any further advice to troubleshoot or resolve this would be appreciated. I guess I'm going to have to replace them and relocate the to external enclosure unless I can resolve this.

 

[Edhawk]

Have you checked that the System Preferences > Energy Saver options are not causing this issue.

Do you have Put hard disks to sleep when possible, selected as shown in the screenshot below.



If yes, try de-selecting this option and the Enable Power Nap option too.

If you have 'Find my Mac' option enabled you will require the Wake for Ethernet network access option enabled. If not then all of the options on this pane should be de-selected.



See if that helps.

 

[greenfrog5]

Unfortunately, both of those settings are already disabled. My settings don't have a Power Nap option. Is that due to my SMBIOS?

Also, when I boot, my new internal HDD (that generally works) sometimes has a warning that "this drive has not been initialized" and I cannot access it. Rebooting (sometimes a few times) resolves this, without any other changes. It seems that something relating to the file system or partition table is not being loaded correctly at startup.

I have not determined any pattern to this, it appears to be random.

 

[Edhawk]

You said the issue arises after 30-minutes, isn't that a pattern? If it happens like that on a regular pattern, then is is likely to be a power/HDD setting kicking in after a set period of time.

Post a screenshot of your Systemwide power settings from Terminal using the following command:

sudo pmset -g

What does it show?

Below is an example of the SystemWide Power settings used with my iMac1 system in my signature.

Systemwide power settings

 

[greenfrog5]

I agree, it seems to have a pattern (now, though in the past it was not as regular/obvious). Sometimes it happens almost immediately after logging in too.

System-wide power settings:
Currently in use:
hibernatemode 0
sleep 0 (sleep prevented by mds, UserEventAgent, mds_stores, UserEventAgent)
Sleep On Power Button 1
ttyskeepawake 1
hibernatefile /var/vm/sleepimage
autorestart 0
disksleep 0
displaysleep 60

 

[greenfrog5]

I just swapped back to my old SSD w/ High Sierra (clean install without many apps or settings). I've been running for an hour, and error -50 has not occurred on the HDD that typically does.

Currently, this is booted with a slightly different Clover EFI. And, an hour is not THAT long. But I do expect the error -50 seems to be Catalina-specific and will probably not occur on this High Sierra install. We'll see whether this HS install experiences freezes too... And I'll compare/text the Clover EFIs to see what significant differences there may be.

 

[Edhawk]

It might be OS/APFS related, I doubt the issue is EFI related.

How is the HDD formatted, GUID/GPT is expected but HFS+ or APFS?

if you can remember, which version of macOS was used to format the drive, High Sierra or a newer OS?

I know you have the drive split into various partitions, are they all formatted the same but used for different storage purposes? Or do you have a mix of NTFS, ExFat, Fat32, HFS+, APFS and ext4 formatted partitions on the drive?

Are you using Paragon NTFS for Mac, or any other similar applications? As these applications may cause the HDD to unmount individual partitions on the drive. Depending on how you have them setup.