Sidebar Sidebar
Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.

Resolved RAM Usage, Swap Files, SSD Degradation...

D

Droid13

macrumors 6502
Original poster
Jul 22, 2009
315
106
United Kingdom
There have been a lot of posts on the above subject scattered throughout the Apple Silicon, MacBook Air and MacBook Pro fora and so I decided to delve deeper into the close to 9-year-old 512GB SSD of my heavily used MacBook Pro that has 16GB RAM.

So far I have established that at a typical time when I am working (Safari with twelve tabs open, one VMWare Horizon session, Microsoft Teams, Microsoft Outlook, Apple Mail, Apple Music) I get the following memory stats from iSTAT:

PRESSURE
Memory Pressure 25%
Memory 62%
App Memory 6.97GB
Wired 2.40GB
Compressed 1.63GB
Cache 4.90GB

MEMORY
Wired 2.40GB
Active 5.94GB
Compressed 1.63GB
Free 6.03GB

SWAP MEMORY
0KB of 0KB

What can I do to formally assess SSD status beyond SMART status and "I don't think there is a problem"?

I suspect my SSD will not have experienced huge amounts of Read/Write activity given there doesn't appear to be any swapping going on.
 
This is how I reduce the problem of "excessive swapping" on drives:

1. TURN OFF virtual memory disk swapping using terminal. I've done it on my Macs for a while now, works great, and NEVER have had a memory-related crash.
However, you must have sufficient RAM for the OS to "live in the physical RAM", AND, you must take care not to overload the physical RAM with too many apps or tabs open (I don't use tabbed browsing ... AT ALL).

2. TURN OFF Spotlight indexing. I've done this on EVERY Mac I use since Spotlight was introduced years ago. I don't want constant indexing or SL messing with my drives. Works fine for me. I have other ways to search when needed.

Just checked using activity monitor:
Activty Report.jpg
 
  • Haha
  • Like
Reactions: eas, LeeW and rocketbuc
Fishrrman said:
This is how I reduce the problem of "excessive swapping" on drives:

1. TURN OFF virtual memory disk swapping using terminal. I've done it on my Macs for a while now, works great, and NEVER have had a memory-related crash.
However, you must have sufficient RAM for the OS to "live in the physical RAM", AND, you must take care not to overload the physical RAM with too many apps or tabs open (I don't use tabbed browsing ... AT ALL).

2. TURN OFF Spotlight indexing. I've done this on EVERY Mac I use since Spotlight was introduced years ago. I don't want constant indexing or SL messing with my drives. Works fine for me. I have other ways to search when needed.

Just checked using activity monitor:
View attachment 2035620
Click to expand...

I saw your advice on one of the other threads and I think I can give it a go and see - from what I can see I need about 10GB for my OS / what I do and I have 16GB. Having said that with my current setup I seem to have squeezed close to nine years from this SSD.

What I was wondering is whether there is a tool that I can use to actually assess "age" or wear on the SSD? I am curious to see the impact of different ways of doing things.
 
there is nothing to gain to turn out swap - all engineers on whatever platform will tell and show you that almost all pagefile operations are small random reads or larger sequential writes, both of which are types of operations that SSDs handle well.

Yes, there is a problem when you are confronted with a bug resulting in excessive writing to the SSD - but “don't swap if you can help it", is misleading, at least for MacOS (as well as Windows and for Linux probably too). MacOS, particular recent versions, will always try to fill up RAM with cached data that it thinks is going to be needed quickly, and will deliberately swap other stuff to disk. It does this irrespective of how much RAM you have. Disabling swapping is a bad idea too, because some programs can require huge amounts of memory reserved for them (think e.g. Photoshop), and you can easily get high memory pressure and "out of memory" messages. It depends on usage, but swapping is always useful to have for extreme situations.
 
  • Like
Reactions: eas, chabig, kitKAC and 1 other person
jz0309 said:
Your best peace of mind will be to get AC+, that is if you’re still eligible…
Click to expand...

Oh we are coming up on nine years down the line with no demonstrable problems with the SSD. I just wanted to know if there was any software that could quantify SSD wear is all.
 
  • Like
Reactions: jz0309
Droid13 said:
Oh we are coming up on nine years down the line with no demonstrable problems with the SSD. I just wanted to know if there was any software that could quantify SSD wear is all.
Click to expand...
9 years is a good run ... just make sure you have backups :)
 
Droid13 said:
What can I do to formally assess SSD status beyond SMART status and "I don't think there is a problem"?
Click to expand...

Not sure what smart status you are looking at. DriveDX might give you what you are looking for. An aging HD shows:

Screen Shot 2022-07-31 at 7.37.10 PM.png

for my new boot SSD DriveDX give some additional information:

Screen Shot 2022-07-31 at 7.40.11 PM.png
 
Droid13 said:
I just wanted to know if there was any software that could quantify SSD wear is all.
Click to expand...
HDFan said:
Not sure what smart status you are looking at. DriveDX might give you what you are looking for.
Click to expand...
I too recommend DriveDx:

DriveDx - the most advanced drive health diagnostics and monitoring utility

DriveDx - the most advanced drive health (S.M.A.R.T.) diagnostics and monitoring utility. Save yourself the data loss and downtime that is associated with unexpected SSD and HDD failures. Don't worry about losing your important data, music, and photographs.
binaryfruit.com binaryfruit.com

• Allows trial period
• One-time fee/cost — a huge perk nowadays
• Cost is very affordable, in my opinion

@jav6454 /\

HDFan said:
for my new boot SSD DriveDX give some additional information:

View attachment 2037314
Click to expand...
By “new” you mean you recently made it a boot drive? Or is the drive “new” as in low age?

I realize even with plenty of RAM some workflows, such as video editing/production, require a lot of storage I/O. However, that usage looks mighty intimidating to my M1’s 8.2 TBW.
😯

:)
 
Last edited:
HDFan said:
Not sure what smart status you are looking at. DriveDX might give you what you are looking for. An aging HD shows:

View attachment 2037313

for my new boot SSD DriveDX give some additional information:

View attachment 2037314
Click to expand...

I have run DriveDx on the close to nine-year-old SSD in my MBP. The overall health and performance ratings are at 100%.

Other notes with percent remaining where available:

Wear Levelling Count - 87%
Host Reads - 37.7TB (99%)
Host Writes - 55TB (99%)

Power On Hours - 6,759 (98%)
Power Cycle Count is 21,680 (78%)
Unsafe Shutdown Count - 1,096 (99%)

Everything else is good; no adverse findings.
 
  • Like
Reactions: MacCheetah3
rpmurray said:
Almost one unsafe shutdown for every six hours of usage. How did you manage that?
Click to expand...

I do not know.

There have certainly been moments in the past when I have thought the machine had died (cyclical restarting and crashing) but that was rare and short-lived.

More recently the battery life has been poor (79% health, real-world battery life just under three hours) and it has not infrequently died without warning. Or while in sleep mode en route somewhere.

One unsafe shutdown every six hours is mental.
 
rpmurray said:
Almost one unsafe shutdown for every six hours of usage. How did you manage that?
Click to expand...
Droid13 said:
I do not know.

There have certainly been moments in the past when I have thought the machine had died (cyclical restarting and crashing) but that was rare and short-lived.

More recently the battery life has been poor (79% health, real-world battery life just under three hours) and it has not infrequently died without warning. Or while in sleep mode en route somewhere.

One unsafe shutdown every six hours is mental.
Click to expand...
My M1’s SSD supposedly had 11 of these unsafe shutdowns. However, I have no recollection of power outages, crashes, or forced system shutdowns.

Unfortunately, the built-in help was not insightful enough.

DriveDx said:
#13 Unsafe Shutdowns

Contains the number of unsafe shutdowns of the SSD. This count is incremented when a shutdown notification (CC.SHN) is not received by the SSD controller prior to the loss of power.
Click to expand...

Therefore, I went Web searching and found a Reddit thread:

lutusp said:
It probably means shutdowns while mounted, or while data was cached for writing but not committed, something like that. I think it's separate from the issue of overall device health.
Click to expand...
edman007 said:
Maybe, as this is smart info, it's probably referring to poweroff without waiting for the last flush to complete. Basically the shutdown was too quick. I would shutdown a few times and see if it increments, if not I wouldn't worry, if it is you may need to do something to give the drive more time at shutdown.
Click to expand...

With that speculation, I wonder if “[app_name] prevented shut down” or general housekeeping (e.g. Spotlight indexing) could also potentially cause this scenario.

P.S. If you hover/click on the raw value, one of the details is Last Modified (essentially, the most recent occurrence).
— Mine says the last was back in December of 2021.
 
Last edited:
Fishrrman said:
This is how I reduce the problem of "excessive swapping" on drives:

1. TURN OFF virtual memory disk swapping using terminal. I've done it on my Macs for a while now, works great, and NEVER have had a memory-related crash.
However, you must have sufficient RAM for the OS to "live in the physical RAM", AND, you must take care not to overload the physical RAM with too many apps or tabs open (I don't use tabbed browsing ... AT ALL).
Click to expand...
I'd be careful about this. Some apps can behave weird when there is no swap file. For light usage, I find the swap file is never/rarely touched anyways (since the introduction of memory compression in mavericks). The system will crash if there is even a temporary spike in RAM usage if there is no swap file. For the lolz I disabled the swap file and the system crashed at ~20gb with an out of memory error even though I have 32 GB.

If the internal SSD is large enough, chances are, even with moderate swap usage, it wont become an issue. Just make sure it's not excessive (e.g. swapping 10+gb constantly) Even old 2013 era 256gb SSDs could do 2000 TB writes before it dies. Those SSDs were warrantied for around ~75 TB of write endurance. The intel SSD was just 15TB.

Today's off the shelf SSDs are warrantied for 300 TB for 500gb, 600TB for 1TB and 1200 TB for 2TB. Apple's SSDs are likely even better.

My SSD usage ranges from 20 to 50gb per day. Even at a whopping 100 gb per day every day, that's 8 to 32 years to hit the warranty amount of a 500gb - 2TB SSD.
 
saudor said:
I'd be careful about this.
Click to expand...

saudor said:
My SSD usage ranges from 20 to 50gb per day. Even at a whopping 100 gb per day every day, that's 8 to 32 years to hit the warranty amount of a 500gb - 2TB SSD.
Click to expand...

That's my take-home. I haven't played with how MacOS uses swap files and I have shown myself that the way I use my machine, the SSD could last anything up to another 50-odd years (assuming I don't kill it with whatever is causing unsafe shut-downs).
 
saudor wrote:
"I'd be careful about this. Some apps can behave weird when there is no swap file."

All I can report is my own experiences after doing so (disabling VM).
I have NEVER had a memory-related crash, ever.
 
You must log in or register to reply here.
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.
Top Bottom
Back