2019 iMac 27" 2TB SSD in boot loop after trying to update to Sequoia 15.8; occasionally shows startup disk but most actions crash it--can you help us end the cycle and clean reinstall?

Screaming Panda wrote:

I was able to create a new volume on it to install Sonoma. But neither First Aid nor Erase have worked on the startup disk regardless of whether we were in Internet Recovery or booting from some other recovery mode or the boot disk I made with Sonoma.

Were you able to install Sonoma successfully on the new APFS volume? And does that new Sonoma install boot? More details on this installation or attempted installation will help provide more clues.

As for First Aid, even if the First Aid summary says everything is "Ok", click "Show Details" and scroll through the report to see if any unfixed errors or warnings are listed. If there are any, then run First Aid again until they are gone. If after several scans the errors/warnings remain, then they cannot be fixed. How to deal with them depends on where those errors/warnings are located (some may be able to be ignored, others will require erasing the drive....how depends on if only the Data volume is affect, or the hidden Container.

Also, you need to run First Aid on the hidden Container. Unfortunately Disk Utility hides the hidden Container by default so you need to click "View" within Disk Utility and select "Show All Devices" before the hidden Container and physical drives appear on the left pane of Disk Utility. You may also want to run First Aid on the physical drive as well.

Try installing macOS onto an external USB3/4 SSD. Just make sure to erase the whole physical external drive (will destroy all data on that drive) as GUID partition and APFS (top option). Option Boot to select the external drive for booting if necessary (sometimes this may be necessary during phase 2 of the install process even when installing to a 2nd volume on the internal SSD). Except for the most extreme internal drive failures, this option should work to verify the rest of the iMac is functioning correctly.

The first RAM slot is not working. I believe I may have bent a connector--argh!--when first trying to see if the RAM was the culprit. Not sure. All other RAM slots are fine. As a result I'm only using slots 2 and 4 for now. My partner says the behavior of the computer has been identical before and after I tried testing each RAM stick individually and doubts that has a causal relation to the boot loop. Could the fact that it won't boot now when the first slot is full be a logic board problem and connected to boot looping, or no?

Certainly possible. If the bent contact is touching another pin or any other metal area, then it may be shorting out causing odd behavior.

Here's the weird thing. If we leave it off overnight, it almost always boots in some way that recognizes the SDD. In Recovery Mode it shows the SSD. It hangs on First Aid, tho. And the erase button hasn't worked. It's initiated, but then hung. The fact that it sort of heals itself for lack of a better term if left alone and rebooted in the mornings makes me think it's not a hardware failure.

May point to a heat related issue with the SSD. The critical item is whether you are referring just to the APFS volume(s) being seen or the physical SSD. If the physical SSD is not appearing properly unless after having the system off for a while would indicate the SSD may be overheating & failing. Remember that Disk Utility hides the physical drives from view by default so use the information I provided above to enable the showing of the physical drives & hidden Containers.

FYI, if you can boot to an external drive, then there are ways of checking the health information of the internal SSD. Unfortunately these NVMe blade style SSDs have very little health information, but it never hurts to check. Actually you can try running the following command in the Terminal while booted into Recovery Mode to see if it will report any health information for the internal SSD (of course the physical internal SSD must be visible in Disk Utility and you will need to verify the device identifier for the physical SSD which normally would be "disk0", but with the issues your iMac is having, it could be something else).

I'm using device identifier "disk0" in this command, but if your SSD is using a different identifier, then replace "disk0" with the correct device identifier for your current boot. Also, it is best to copy & paste this command because there are a lot of odd options & spacing is critical. You can edit the pasted line by using the left arrow to move to the right side of the zero in "disk0" to delete the zero and replace it with the number one. Press the "Return" key to execute the command.

diskutil info -plist disk0 | plutil -extract "SMARTDeviceSpecificKeysMayVaryNotGuaranteed" json -r - -o - | grep -v '_1"'

Screaming Panda wrote:

This past year it had slowed down a bit with Sequoia updates. I chalked this up to programmed obsolescence.

Scores of people are running Sequoia on those machines. Degraded performance could be due to conflicts from incompatible installed items, or due to SSD failure.

it would still crash eventually and boot loop again. Or just shut itself down. So I switched to trying to assess and fix things in recovery mode. It seems successive crashes obliterate the ability to start up normally.

This has symptoms that are similar to failed SSDs I have seen. Circumstantial evidence, however.

My hypothesis is the initial crash occurred due to a failed software update to Sequoia 15.8 and the OS is corrupted--rather than a hardware failure--but either way it has been unstable with the SSD appearing and disappearing in recovery mode.

That seems like it could well be due to SSD (or other) failure. SSD appearing and disappearing in recovery mode does not seem like an OS failure (the OS isn't even loaded at that point) but more likely SSD failure or intermittent failure.

The first RAM slot is not working. I believe I may have bent a connector--argh!--when first trying to see if the RAM was the culprit. Not sure. All other RAM slots are fine. As a result I'm only using slots 2 and 4 for now.

This makes it more complicated as you have other hardware problems there as well.

My partner says the behavior of the computer has been identical before and after I tried testing each RAM stick individually and doubts that has a causal relation to the boot loop. Could the fact that it won't boot now when the first slot is full be a logic board problem and connected to boot looping, or no?

It could be a logic board problem, but that's speculation.

Here's the weird thing. If we leave it off overnight, it almost always boots in some way that recognizes the SDD.

If solder joints or integrated circuitry have degraded, they might fail when warm, then come together for a better electrical connection when cold. Or vice versa. Hence different behaviors can be seen on cold boot versus warm boot. I have worked on spacecraft devices and hardware for decades and this is why we always test them over a wide range of operating temperatures before launch. This will induce failure modes not seen at some other temperatures. This is also indicative of a hardware issue with your computer.

In Recovery Mode it shows the SSD. It hangs on First Aid, tho.

Another symptom consistent with hardware failure.

And the erase button hasn't worked. It's initiated, but then hung.

The SSD is not responding properly.

The fact that it sort of heals itself for lack of a better term if left alone and rebooted in the mornings makes me think it's not a hardware failure.

That makes me think IT IS hardware failure. Hardware and tiny integrated circuitry and solder joints behave differently under different temperatures. See my comments on that above. Software does not change with temperature but hardware does. Wires and metal literally expand and contract with changes in temperature.

I took it to the local Apple Store in NYC. They're running advanced diagnostics. I'll report back what they find.

That will be more definitive.

I still believe the iMac is recoverable as a machine and does not have a hardware failure since it seemed to grind to a halt due to the OS, not some typical mechanical/disk failure type of situation.

I think the opposite based on what you have shared: hardware failure.

I am guessing that this iMac needs to be replaced.

As a "hail Mary," have you tried the following:

What to do before you sell, give away, trade in, or recycle your Mac - Apple Support

That would completely wipe the Mac and restore a plain macOS from Apple's servers. This might lead you to restore your Mac for use, assuming its hardware is ok. You would need a good backup of your files, which it sounds like you will need anyway.

P.S. Do you have a Fusion drive? They have proven to be problematic and more prone to failure than either SSDs alone or HDDs alone.

If the disk file catalog is so deeply corrupted that in Recovery, Disk Utility can't even run on it, then the link I provided for wiping the machine to return to the factory setup will probably work. Because those would be soft errors.

If it is hardware, it will just continue to get worse. These things don't self heal, but they can behave differently at different temperatures. Blowing cold air on the computer externally won't really affect it; the internal temperatures can be 35°C to 50°C (95°F to 122°F) during inactivity, and as much as 65°C to 95°C (150°F to 200°F) when being worked intensively.

A 7-year old iMac can fail at any time. Not pushing it "rigorously" does not matter, look at those temperatures above, cycling from cool to hot to cool to hot, tens of thousands of times over 7 years can weaken or degrade circuitry or solder joints, and any of the electronics can fail, SSD, logic board, other components. Most people don't push their computers very hard, people that do intensive video editing do, but running at a constant hot temperature is actually better for the electronics than many temperature cycles.

I've seen everything: I have a 2010 MacBook Air still working on all original parts (including its battery which has 81% original capacity); a 2013 MacBook Air had its SSD and battery fail after 7 years, and it is still running well after replacing both the battery and SSD. I had a 2008 iMac that was still running after 13 years until its power supply failed in 2021, at which point I recycled it. I am writing this right now on a 2015 iMac which is running well (replaced the internal HDD with an SSD since the HDD was too slow). I also have newer (2025) Macs in use. What I am saying is that it is not unusual for 7-yr old computer to fail, but some will last longer. My employer requires all PCs and Macs to be "replenished" after 3 years, that's their policy.

I'll be interested to hear what the Apple technicians conclude after testing this iMac. Again, if the disk structure is so corrupted that it just freezes the Mac whenever accessed, then the complete erase/format and fresh OS install should restore it to working condition. But if that were the problem, it wouldn't improve overnight -- that behavior makes, in my view, hardware more likely the cause. But you will find out soon! Let us know.

OK. Perhaps I'm living on cope lol. We have frequent time machine backups leading up to the crash. It's a 2TB SSD, not Fusion. It was only ever about half full at maximum.

If solder joints or integrated circuitry have degraded, they might fail when warm, then come together for a better electrical connection when cold. Or vice versa. Hence different behaviors can be seen on cold boot versus warm boot.

Yeah, this part makes me think it's way more likely hardware. I've never had an SSD fail, so perhaps this is just the straw that broke the camel's back. But I still do wonder if it's just corrupted data. Since all Diagnostics tests have shown a 100% healthy system up until now. And I encountered the below:

"Severe Software Corruption: Apple Diagnostics only looks for hardware faults. If your macOS partition or file catalog structure is deeply corrupted, Disk Utility First Aid will freeze trying to repair it, but diagnostics will report no hardware issues."

To be fair, I'm assuming the restarts in the morning differ due to temperature, but have not proven that. We've put it near an AC unit to keep it as cold as possible just in case and it still didn't seem to help, so the heat and thus hardware being the culprit may or may not be triggering the failures.

But I'm not ruling it out. It's just not a machine that's been put through super rigorous use. Basic graphic design work. I've run machines for video editing, effects processing, tons of transcoding and writing of massing 100 GB files. Not this one. 5.5 years on a native Apple internal SSD failing amid a simple update? Eventually, sure, the reaper comes for us all.

Thanks for your candid feedback!