MP 1,1-5,1 Boot SSD periodically unmounting/remounting?

[Valdaquendë]

The last two mornings, I've come into the office to find a MacOS notification admonishing me to properly eject my boot disk <Mac_SSD> before removing it. If seems, then, that for some unexplained reason, my boot SSD appears to be un-mounting itself in the middle of the night and then remounting. If this is truly happening, the episode must be quite momentary, as the system is still running (has not shut down or restarted), no apps appear to be affected and I see no sign of other irregularity. My CPU temps vary between 50˚C. and 65˚C. and all other temps appear well within normal limits. Probably at the lower end of that range when this happens as the system is not actually being used at that time.

I have run Tech Tool Pro to verify the SSD, including S.M.A.R.T. tests and surface scans and have tested the other system components; all components appear to have passed with flying colors. I tried checking the logs in the System Report but have no idea what, in all those entries, I should be looking for. Naturally, this is a concern; has anyone else seen this sort of behavior?

This brings up a tangential question. I checked my SSD's firmware version; it appears to be the latest release. I have been wondering if tsialex's advice that the Mac Pro's BootROM should be flashed periodically is an idea that is, perhaps, applicable to SSDs as well.

NOTE: I should point out, here, that two nights ago I formatted a 2T SSD and cloned this one to it (using CCC), as part of another project. So I have a clean copy of the drive. And, though I did properly eject it afterward, the nonsensical notion that the OS might be referring to that one did cross my mind.

Any advice will be gratefully received and carefully considered. Thanks in advance for the time you spent reading and considering this.

Valdaquendë
==========================
System Info: Mac Pro (5,1 Mid-2010)
CPU - 3.33Ghz 6-core (Westmere)
RAM - 16GB 1333MHz DDR3
GPU - Sapphire Radeon Rx 580 PULSE
OS - MacOS 10.12.6
Boot Drive - 1T Crucial MX500 SSD (CT1000MX500SSD1)
Data Drives - 3 HGST 4TB NAS drives

 

[tsialex]

The last two mornings, I've come into the office to find a MacOS notification admonishing me to properly eject my boot disk <Mac_SSD> before removing it. If seems, then, that for some unexplained reason, my boot SSD appears to be un-mounting itself in the middle of the night and then remounting. If this is truly happening, the episode must be quite momentary, as the system is still running (has not shut down or restarted), no apps appear to be affected and I see no sign of other irregularity. My CPU temps vary between 50˚C. and 65˚C. and all other temps appear well within normal limits. Probably at the lower end of that range when this happens as the system is not actually being used at that time.

I have run Tech Tool Pro to verify the SSD, including S.M.A.R.T. tests and surface scans and have tested the other system components; all components appear to have passed with flying colors. I tried checking the logs in the System Report but have no idea what, in all those entries, I should be looking for. Naturally, this is a concern; has anyone else seen this sort of behavior?

This brings up a tangential question. I checked my SSD's firmware version; it appears to be the latest release. I have been wondering if tsialex's advice that the Mac Pro's BootROM should be flashed periodically is an idea that is, perhaps, applicable to SSDs as well.

Any advice will be gratefully received and carefully considered. Thanks in advance for the time you spent reading and considering this.

Valdaquendë
==========================
System Info: Mac Pro (5,1 Mid-2010)
CPU - 3.33Ghz 6-core (Westmere)
RAM - 16GB 1333MHz DDR3
GPU - Sapphire Radeon Rx 580 PULSE
OS - MacOS 10.12.6
Boot Drive - 1T Crucial MX500 SSD (CT1000MX500SSD1)
Data Drives - 3 HGST 4TB NAS drives

Did you tried DriveDX to see if you can get any clues?

I have a Crucial BX500 that did exactly that and I couldn't continue use it on Mac, it's on my bench PC now.

 

[Valdaquendë]

Thanks, I'll try that right away.

Do you think there is any benefit to reinstalling an SSD's firmware with the idea that this may reset the controller or other parameters? In reading an in-depth discussion of NvMe drives by an Intel engineer, he mentioned that the controllers are somewhat of an Achilles heel.

 

[tsialex]

Thanks, I'll try that right away.

Do you think there is any benefit to reinstalling an SSD's firmware with the idea that this may reset the controller or other parameters? In reading an in-depth discussion of NvMe drives by an Intel engineer, he mentioned that the controllers are somewhat of an Achilles heel.

Most flashers don't accept re-flashes of the same version or downgrades, so, unless you hack the firmware validation, you not gonna be able do it.

 

[Valdaquendë]

Good to know. I ran DriveDX; the results came back clean; no errors, 60% life left, Overall Performance Rating 100%, etc. I'll attach the report, on the off-chance you care to look at it. The only matter of concern that I could see was that the device had been over temperature for a total of 37 minutes during its tenure. Having seen that, I used Macs Fan Control to set the fans to keep the system below 65˚C. I'm assuming the device is basically OK but am going to keep a 2T spare regularly synce'd just in case.

I am curious, though, as to the validity of the principle of periodically re-flashing and will try to find a manufacturer's venue for asking that question. I'll report any results.

I had not used DriveDX, though I'd heard of it; a spectacularly good tool; I'm going to get a consultant's license for it and add to my diagnostic/utility toolbox along with DiskDrill, TTP, Cocktail, Macs Fan Control, CCC, EtreCheck and Onyx. I'd love to know about any others you think I am missing.

I hurt my back yesterday, for the first time in two years (I have three fused vertibrae at the base of my spine), so I am desk/bench/couch-bound today. So I'm going to practice my soldering skills. I am getting ready to overhaul the 12-core I got in November (the one with the formerly-bent case). That's the first SPI I'll want to replace. And the 2TB spare SSD I'm keeping sync'ed will be the boot drive for that system, which will replace my current office system.

Have a great weekend!

 

[Valdaquendë]

Did you tried DriveDX to see if you can get any clues?

I have a Crucial BX500 that did exactly that and I couldn't continue use it on Mac, it's on my bench PC now.

My first SSD was a BX500 but then I found an article in which they had tested the throughput between the BX and MX models and found that the MX's real-world performance was over twice that of the BX model line. That was a few years ago and I don't recall where I saw it. I did some checking today but all articles I could find point to fairly close performance between the two

 

[tsialex]

My first SSD was a BX500 but then I found an article in which they had tested the throughput between the BX and MX models and found that the MX's real-world performance was over twice that of the BX model line. That was a few years ago and I don't recall where I saw it. I did some checking today but all articles I could find point to fairly close performance between the two

It's a DRAM-less low endurance extreme entry level drive, I never should have installed it on my MBP. It's working fine where I'm using it now, but I'll never buy one again.

 

[tsialex]

I hurt my back yesterday, for the first time in two years (I have three fused vertibrae at the base of my spine), so I am desk/bench/couch-bound today.

Back when I was in the university, I had a car accident and broke my left femur in three pieces. I have more screws that I can count, two support plates and I had to remove both left and right hip tips for bone grafts, so I clearly understand bad/debilitating days - had one last week. I hope you get better.

So I'm going to practice my soldering skills. I am getting ready to overhaul the 12-core I got in November (the one with the formerly-bent case). That's the first SPI I'll want to replace. And the 2TB spare SSD I'm keeping sync'ed will be the boot drive for that system, which will replace my current office system.

Good Luck!

Have a great weekend!

To you too!

 

[skodises]

DriveDX is without a doubt one of the better tools you can have in your cMP toolbox- it has already saved me a significant amount of grief, as I have an absurd amount of storage online (with different physical drives for different client activites, kept rigidly separated in case of discovery actions in the legal arena). As soon as a drive falls below ~75% health, I scrap it.

However, there's still a cautionary note here. It is not a panacea. There are failure modes in both spinners and SSDs that DriveDX can't anticipate. As an example, I bought a 1TB Kingston NV1 NVMe drive this past November, to use on my 7101A as a cheapish fastish backup drive for one particular set of files that I'm currently actively modifying. I heatsinked the heck out of it, and it never ran above 65degC- somewhat cooler than my usual Samsung 970 EVOs. About a week ago, my machine abruptly hung and crashed midday (which essentially never happens, due to my very conservative usage style). When it came back up, the Kingston drive was simply absent: and no amount of tweeking, reseating, or sacrificing of chickens could bring it back to life. It died spectacularly after only 4 months of use as a backup drive: it went from DriveDX reporting 100% health to freakin' deceased in moments.

C'est la guerre. I had fun putting it to the hammer test, as I do to destroy all my retired drives. And that will be the last time I try to "save money" on anything intended to hold valuable data. Hope that helps someone: back up early, back up often, and be careful out there!