Caution! Troubles with 870 evo drives on cMP 5,1 !

[tsialex]

Please name one where that's a verified issue if you don't mind. Oh it was considered the bogeyman before anyone understood what it really did but that was dismissed years ago.

You are mixing different things. I'm not talking about mkt issues from 6~8 years ago. SSD firmwares lie all the time, the most recent discovered issue is with write cache behavior.

Anyway, people have been testing this for a long time and OpenCore devs implemented lot's of quirks for drives that have problems with TRIM. Please take a look at the commits and SetApfsTrimTimeout:

1. SetApfsTrimTimeout
   Type: plist integer
   Failsafe: -1
   Requirement: 10.14 (not required for older)
   Description: Set trim timeout in microseconds for APFS filesystems on SSDs.
   The APFS filesystem is designed in a way that the space controlled via the spaceman structure is either used or free. This may be different in other filesystems where the areas can be marked as used, free, and unmapped. All free space is trimmed (unmapped/deallocated) at macOS startup. The trimming procedure for NVMe drives happens in LBA ranges due to the nature of the DSM command with up to 256 ranges per command. The more fragmented the memory on the drive is, the more commands are necessary to trim all the free space.
   Depending on the SSD controller and the level of drive fragmenation, the trim procedure may take a considerable amount of time, causing noticeable boot slowdown. The APFS driver explicitly ignores previously unmapped areas and repeatedly trims them on boot. To mitigate against such boot slowdowns, the macOS driver introduced a timeout (9.999999 seconds) that stops the trim operation when not finished in time.
   On several controllers, such as Samsung, where the deallocation process is relatively slow, this timeout can be reached very quickly. Essentially, it means that the level of fragmentation is high, thus macOS will attempt to trim the same lower blocks that have previously been deallocated, but never have enough time to deallocate higher blocks. The outcome is that trimming on such SSDs will be non-functional soon after installation, resulting in additional wear on the flash.
   One way to workaround the problem is to increase the timeout to an extremely high value, which at the cost of slow boot times (extra minutes) will ensure that all the blocks are trimmed. Setting this option to a high value, such as 4294967295 (a.k.a. -1) ensures that all blocks are trimmed. Alternatively, use over-provisioning, if supported, or create a dedicated unmapped partition where the reserve blocks can be found by the controller. Conversely, the trim operation can be mostly disabled by setting a very low timeout value, while 0 entirely disables it. Refer to this article for details.
   On macOS 12+, it is no longer possible to specify trim timeout for APFS filesystems. However, it can be disabled by setting 0.
   Note: System Information may still report TRIM Support: Yes with this option set to 0. Yet, this is false positive and trim is de facto disabled.

 

[mikehalloran]

has anyone tested a 3,5 to 2,5 adapter with a pcb like this?
View attachment 1986047

Did anyone test to disable trim?

That PCB is merely a pin extender. I've lost count of the number of SSDs I've installed using something similar (over 500, less than 1000) but this style promotes better cooling.

TRIM is not an issue. The MacOS blocks it automatically in a 2.5" SSD unless you unblock it. It's never blocked in a blade type SSD.

 

[czs]

FWIW, just installed a brand-new 1TB 870EVO into the drive bay of my cMP5,1 (2012) and this has the not-seen-on-warm-reboot issue.

Tried using both Apple EFI & OpenCore + every OS from HS to Monterey - no dice - the drive only appears on power-on boot. I was pretty sure this is unrelated to the EFI and\or the OS but still had to check just in case. And yes, done several PRAM and SMC resets - again, no dice - only seen at power-on.

Also, while some have apparently had good luck solving this with a firmware update, I can confirm that is not a reliable solution - why? Because my drive has the latest firmware as of the date of this post (SVT02B6Q).

I know many say just go with the 860 but yes, I'm very stubborn so not quite ready to toss in the towel...

 

[dlopan]

FWIW, just installed a brand-new 1TB 870EVO into the drive bay of my cMP5,1 (2012) and this has the not-seen-on-warm-reboot issue.

Tried using both Apple EFI & OpenCore + every OS from HS to Monterey - no dice - the drive only appears on power-on boot. I was pretty sure this is unrelated to the EFI and\or the OS but still had to check just in case. And yes, done several PRAM and SMC resets - again, no dice - only seen at power-on.

Also, while some have apparently had good luck solving this with a firmware update, I can confirm that is not a reliable solution - why? Because my drive has the latest firmware as of the date of this post (SVT02B6Q).

I know many say just go with the 860 but yes, I'm very stubborn so not quite ready to toss in the towel...

Try a hail mary and reinstall the latest firmware if it lets you. Just for grins

 

[czs]

Try a hail mary and reinstall the latest firmware if it lets you. Just for grins

Hi @dlopan - thanks for the suggestion. Already have 144.0.0.0 as it's needed it to get beyond Mojave (or maybe HiSi - forget which installs this)...

Just realized you probably meant the drive firmware - not the Mac. I do have the current version so perhaps you meant re-install?

 

[avro707]

Try a hail mary and reinstall the latest firmware if it lets you. Just for grins

I think this is worth a try on the SSD. Try upgrading the firmware again just in case.

I have two of these 870 drives and neither gave any issues.

 

[czs]

I think this is worth a try on the SSD. Try upgrading the firmware again just in case.

I have two of these 870 drives and neither gave any issues.

Hi @avro707 - good info. Let me please confirm - you have your EVO 870s installed into a cMP5,1 (2012) and connected via the backplane (not a PCIe card) and able to see them on a warm-reboot?

If so, can you please post your SSD firmware rev?

 

[dlopan]

Hi @dlopan - thanks for the suggestion. Already have 144.0.0.0 as it's needed it to get beyond Mojave (or maybe HiSi - forget which installs this)...

Just realized you probably meant the drive firmware - not the Mac. I do have the current version so perhaps you meant re-install?

Yes. Sorry.

 

[czs]

Yes. Sorry.

No worries. I did some digging on this and thought there was a glimmer of hope when I saw both a Windows and Mac edition of the firmware up on the Samsung site. Both were the same version so I figured Samsung had made some platform specific tweaks for the Mac but alas, not so. The delta was actually just how the FW is written to the USB drive. And of course, when trying the install, the SSD says it is already current.

Back to digging...

 

[dlopan]

You just have to have the right way to find them. No truffles in NM.

 

[Macschrauber]

Have you compared the firmware files?

Terminal: diff file1.bin file2.bin

no output if files match exactly

 

[czs]

Have you compared the firmware files?

Terminal: diff file1.bin file2.bin

no output if files match exactly

The files are definitely different but because the firmware versions are the same, I presume the only difference is how the bootable installer is built - one for Mac and the other for Wintel. Perhaps they also made changes to the firmware but that wouldn't make much sense especially without changing the firmware version.

 

[Macschrauber]

I think of the firmware binary file.

If you download both updaters, extract and compare the files we know if there are Mac and PC firmwares, even when they give both versions the same number.

 

[zedex]

Hi @csz - I'm interested in your experience because it's pretty close to baseline and I'm still keen to determine whether someone following a strict baseline setup approach might still experience the 'invisible warm boot' issue.

As you mentioned, you've got an 870 installed to the SATA II backplane of your 2012 5,1.

The next 'baseline' protocol would be formatting the 870 SSD as HFS+ with GUID partition map.

Would the new disk [870] show on warm boot after all conventional SATA standards/AHCI protocols are correctly observed..?

This would conclusively prove that APFS is the culprit which is what I've long suspected. Then you could implement one of 2 straight forward workarounds to solve the problem permanently.

+++
Also - if this hypothesis proves incorrect, you could then - with a high degree of confidence - conclude that, this seemingly confounding issue is identical in form and substance to the 'high capacity HDD warm boot issue' written about elsewhere on this forum AND we would all finally have the unambiguous answer to the question...... what is the problem and when does it occur

It would look something like this:

If you own (or have recently purchased a Samsung 870 SSD) you should know that high capacity models (say 2TB or above) will reliably demonstrate the 'invisible warm boot' phenomenon [as described in more detail below].

This problem was originally uncovered and resolved on the 'high capacity HDD warm boot issue' thread on this forum and relates to a small change/update in the AHCI protocol circa 2015 wherein the third trace of the 15-pin SATA power connector was repurposed to signal for 'power saving mode' on large capacity storage device at the behest of industry blah blah blah, etc

This is how you fix it.. [publish solution from the 'high capacity HDD warm boot issue' thread]

FOR THE AVOIDANCE OF DOUBT
- this is not a problem with your SSD firmware
- this is not a problem with 870 models with a capacity of 500GB or 1TB
- this is not a problem with the APFS file system format
- this is not a problem with macOS
- this is not a problem with Samsung quality assurance
- this is not a problem for 7,1 owners (with native SATA III interface and 2019 AHCI protocol compliance)

at this point the confusion should finally stop.

Samsung 870 EVO 500GB owners will reliably expect and confirm that their disk works OOTB.
Samsung 870 EVO 4TB owners will reliably expect and confirm that their disk never shows on warm reboot.
and no one will waste time looking for firmware updates or returning "defective" SSDs to Samsung

 

[Alpha Centauri]

Not a MP but mid 2009 MBP. Very recently cloned patched Cat with SuperDuper from the main 850Evo 1TB to a new 870Evo 2TB. The 870 now in main bay, 850 moved to CD drive bay.

Every boot into 870 (APFS formatted) causes upped temps/fans on idle. BUT strangely Memory Pressure immediately drops into yellow. Activity monitor shows no discernible processes at all. To note in the past, had swaps under 1GB under load but memory was always in the green since I had this machine.

Booting back into the 850Evo and the issue disappears entirely. Unusable the way it is and without a solution soon I'll be returning the new drive.

 

[zgagato]

I don't know if it was already mentioned but I had problems with 870 drives in 2012 MacBooks pro, non retina. In some cases it was working ok, in most other it disappeared after 1st reset. Couldn't get the fresh os to boot. I just gave up and keep on using Crucial MX500 drives. I know, the quality is much lower but at least they work.
Something is done in those drives firmware, they just don't work with Macs. The behaviour is the same like you described here.

 

[Alpha Centauri]

I don't know if it was already mentioned but I had problems with 870 drives in 2012 MacBooks pro, non retina. In some cases it was working ok, in most other it disappeared after 1st reset. Couldn't get the fresh os to boot. I just gave up and keep on using Crucial MX500 drives. I know, the quality is much lower but at least they work.
Something is done in those drives firmware, they just don't work with Macs. The behaviour is the same like you described here.

That solidifys my suspicion even more (see post 140) and mine's a 2009 MBP. I thought it maybe my Catalina patch gone bad in the cloning to the new 870, but the memory pressure with nothing running I've never in 13 yrs of ownership come across. The 850 hums along quite nicely though. Thanks for posting your experience with this.

 

[zgagato]

That solidifys my suspicion even more (see post 140) and mine's a 2009 MBP. I thought it maybe my Catalina patch gone bad in the cloning to the new 870, but the memory pressure with nothing running I've never in 13 yrs of ownership come across. The 850 hums along quite nicely though. Thanks for posting your experience with this.

I service Macs for living. Never had a single issue with SATA Samsungs from 830 till 860 series. Evo and Pro when they appeared. In 830 all of them were "pro" and expensive. Not a single failure. They still work in My 5.1 or PS4. Sometimes 870 works in one unit of mbp 2012 but it fails in another identical unit. Its so random that I decided not to bother.

 

[krakman]

Update: Its been several days since my original post and the 2TB EVO continues to survive warm restart.


2tb 870 EVO suddenly survives restart!!!!!!

Well here is bit of unexpected news!!! It never survived a restart only a cold boot!

I moved the 2TB EVO from the SATA B cd drive connector to the A connector.

The only other "new" things I did:

1. Install the 12.5.1 update - which the drive didn't survive the reboot during the install process, I had to manually force shut down several times during the upgrade.
2. Shut down and power off at mains for several days whilst away from the office
3. Add a 1tb 850 evo to the B CD drive connector
4. I'm using OCLP version 0.4.10

The 2TB EVO has the original firmware that it shipped with

I have successfully done 6 warm reboots in a row this evening.

 

[Alpha Centauri]

I service Macs for living. Never had a single issue with SATA Samsungs from 830 till 860 series. Evo and Pro when they appeared. In 830 all of them were "pro" and expensive. Not a single failure. They still work in My 5.1 or PS4. Sometimes 870 works in one unit of mbp 2012 but it fails in another identical unit. Its so random that I decided not to bother.

Yet..here I am with this issue.

Tbh, I cannot identify the problem but to assume the current 870 firmware is simply too much a stretch to support my 2009 model. I'm open to suggestions. The drive being €200 (2 TB), it would be nice to sort it. I'm still in the return window, would overwrite x7 but it would get resold and that never sits right with one's peronal data having "been" on there.

The machine has had a few SSDs since the first install in 2011. In all this time there's been light-heavy memory swaps, CPU throttling...sure. But never have I run into memory pressure into yellow in all the time running 10.6 to 10.11 and now off course 10.15. Never until now with this 870 clone, not to mention it's occurring just at idle too.

 

[Theidiot1989]

The way I got my Samsung 870 to work was by going into restore mode press command r formatting it to the opetion at the vary top think it’s call adsd or something I just press erase in the options come up in the it’s at the top of the list well in disk utility’s. I did a back up of old system and before hand.. i clicked on restore and since my ssd was formatted it came up on the list which enabled me to back up my old ssd which I had saved on a hard drive. I clicked restore after choosing the Samsung ssd witch was renamed from formatting.. and pressed start or restore or whatever.. it all restored onto Samsung 870 evo. It restarted and worked it works everytime I turn the computer on. I also turned trim mode on through command to speed it up you can find directions online..also hade it in a pci slot the more power one not x4.

 

[Theidiot1989]

The way I got my Samsung 870 to work was by going into restore mode press command r formatting it to the opetion at the vary top think it’s call adsd or something I just press erase in the options come up in the it’s at the top of the list well in disk utility’s. I did a back up of old system and before hand.. i clicked on restore and since my ssd was formatted it came up on the list which enabled me to back up my old ssd which I had saved on a hard drive. I clicked restore after choosing the Samsung ssd witch was renamed from formatting.. and pressed start or restore or whatever.. it all restored onto Samsung 870 evo. It restarted and worked it works everytime I turn the computer on. I also turned trim mode on through command to speed it up you can find directions online..also hade it in a pci slot the more power one not x4.

Btw you gotta restart again after it restarts everything and Starts up. Set the start up drive as the Samsung and restart turn off computer again once that’s done and start up again and it will boot up and work good the first time when it restarts from the restore it’s pretty glitchy once you restart though and turn ur system off in on its fine

 

[8364R4]

+1 here:
- MP51/2010, 870EVO 2TB, SATA @ Bay2, Expansion Drive = WarmBoot issue
- same machine with 860EVO 2TB, SATA @ Bay2, Expansion Drive = works fine
- same machine with 850EVO 1TB, SATA @ Bay1, System Drive = works fine

As there are some very skilled people around here I wonder if one may answer to the the following questions:

With Mojave/APFS is it actually risky to enable TRIM with the 860/850 drives connected to the SATA port?
If so, having TRIM disabled, will I risk a quick dead of the drives or they perform less good?
Or would you suggest anyway to run Mojave with HFS and TRIM enabled?

Performance and reliability are the most important criteria with this system.

Thanks a bunch!

 

[MarkC426]

I have had an 840 evo and 860 evo on the sata ports with Trim enabled for many years, with no issues.

 

[8364R4]

In combination with Mojave and APFS?