ssd swap - high usage of Terabytes Written

[Feffe]

With Firefox swap usage seems very good - usually at 0, getting a bit higher only when running Parallels or intensive Anaconda tasks.

I found mds_store wrote 4GB while the Macbook was sleeping (for around 8-9 hours). Is this too high? I think this is Spotlight - maybe it can be disabled while the Mac is sleeping?

 

[Baff]

I found mds_store wrote 4GB while the Macbook was sleeping (for around 8-9 hours). Is this too high? I think this is Spotlight - maybe it can be disabled while the Mac is sleeping?

I wouldn't worry about that at all.

 

[Fred Zed]

Curious why you haven’t updated to current version of macOS?

Is this your way of asking me what version of the OS I have installed?

 

[Mike Boreham]

I am still getting to grips with this topic so grateful for guidance what I should be looking at.

Smartmontools says my M1 MBA (new Nov 24th) has TBW 15.8TB
DriveDX shows 14.4TB
These compute to a little over 5GB/h.
I am OK looking at DriveDx and Smartmontools but don't understand what to look at in Activity Monitor for the same data over the current uptime period. Is the screenshot below (27hr uptime) the one to look at?
Data written is 178GB but the numbers in the "Bytes Written" column add up to much more than that? Even the 178GB number calculates as 6.6GB/h ... ie more than 5GB/h from Smartmontools and DriveDx.
Thanks for any input.

 

[leons]

First off your SSD usage is fine, even at the highest number you have. Secondly, it looks to me that you are also writing to an external drive?

 

[Mike Boreham]

First off your SSD usage is fine, even at the highest number you have. Secondly, it looks to me that you are also writing to an external drive?

Thanks very much. Yes I am using an external so that explains why AM shows bigger numbers. I remember reading that somewhere in this thread now. Is there any way of using AM to monitor this issue if an external drive is involved?

 

[leons]

I'm "guessing" (maybe someone else can confirm) that the "Data Written" number at the bottom reflects the local disk only, and that the numbers in the "Bytes Written" column reflect everything. If that's the case, the marginally higher number (than your usual bytes-written/hour, STILL well within acceptable limits), might just be a small spike in usage over that particular period of time.

 

[Baff]

I'm "guessing" (maybe someone else can confirm) that the "Data Written" number at the bottom reflects the local disk only,

No, yesterday I transferred about 100GB between 2 external drives and it showed up in the bottom right. But, as soon as I unplugged the external drive that I had written to, the data written number dropped.

 

[leons]

No, yesterday I transferred about 100GB between 2 external drives and it showed up in the bottom right. But, as soon as I unplugged the external drive that I had written to, the data written number dropped.

Dropped the amount xferred to the external drive?

 

[Baff]

Dropped the amount xferred to the external drive?

Yes

 

[ambient_light]

Apparently the latest beta 11.3.6 (?) has fixed it for many people? Can anyone confirm this?

Unfortunately, not fixed for me yet. In fact, the best swapping behavior was so far in the 11.3.4, but then it degraded again in later updates. There‘s still a mild improvement though before the swap is crossing the 4gb watermark.

 

[meropenem]

Finally tried Lightroom Classic CC on my M1 Pro and wow. An additional 500gb written in about half an hour. The fact that people try to deny this is an issue is absolutely unbelievable.
On the latest non-Beta MacOS 11.2.3 for those wondering.

 

[chouseworth]

Finally tried Lightroom Classic CC on my M1 Pro and wow. An additional 500gb written in about half an hour. The fact that people try to deny this is an issue is absolutely unbelievable.

I have had the same problem in the LR Classic Develop module. I posted a question here earlier today asking if anyone had been able to find some kind of workaround until there is a fix. Otherwise I may have to go to Capture One at least temporarily.

 

[meropenem]

I have had the same problem in the LR Classic Develop module. I posted a question here earlier today asking if anyone had been able to find some kind of workaround until there is a fix. Otherwise I may have to go to Capture One at least temporarily.

I read that it is better with GPU accel off, so I've turned it off in the meantime. I tried exploring using Luminar AI but that app swaps even worse and is far laggier.

 

[leons]

Unfortunately, not fixed for me yet. In fact, the best swapping behavior was so far in the 11.3.4, but then it degraded again in later updates. There‘s still a mild improvement though before the swap is crossing the 4gb watermark.

Have you made any of the adjustments suggested in this thread?

 

[Maximara]

I'm "guessing" (maybe someone else can confirm) that the "Data Written" number at the bottom reflects the local disk only, and that the numbers in the "Bytes Written" column reflect everything. If that's the case, the marginally higher number (than your usual bytes-written/hour, STILL well within acceptable limits), might just be a small spike in usage over that particular period of time.

Actually as related in a video cited a while back the Data Written on Activity Monitor supposedly represents all data written to all drives and not just what is written to the boot up drive. But the number being shown paint a different picture so I have no idea.

 

[4k78]

Is this your way of asking me what version of the OS I have installed?

What version of the OS do you have installed?

 

[leons]

Finally tried Lightroom Classic CC on my M1 Pro and wow. An additional 500gb written in about half an hour. The fact that people try to deny this is an issue is absolutely unbelievable.
On the latest non-Beta MacOS 11.2.3 for those wondering.

I haven't run into anyone denying it. Some of us (myself included) have found ways to mitigate the issue so it is no longer a problem. I know that some workarounds have been discussed on the Adobe Forum, but it is possible that those using certain software (e.g. Lightroom) won't see a complete "fix" until Adobe ports them to M1.

 

[leons]

Actually as related in a video cited a while back the Data Written on Activity Monitor supposedly represents all data written to all drives and not just what is written to the boot up drive. But the number being shown paint a different picture so I have no idea.

Based on his response, it seems that "Data Written" in Activity Monitor gives the total of ALL connected drives UNTIL the drive is disconnected, then falls back to the number for the host drive only. In any case, the "real" number is easily extracted by tools or terminal commannds.

 

[osplo]

I know that some workarounds have been discussed on the Adobe Forum, but it is possible that those using certain software (e.g. Lightroom) won't see a complete "fix" until Adobe ports them to M1.

I don't have any issues so far with excessive SSD writes, but I admit I have not installed Lightroom Classic yet.

Anyway, I don't really want to run it on my Air. I want it on my future Apple Silicon iMac. Then I really want LR to be Apple Silicon native.

Hopefully we will have soon both (the AS iMac and the native LR Classic).

 

[Maximara]

Based on his response, it seems that "Data Written" in Activity Monitor gives the total of ALL connected drives UNTIL the drive is disconnected, then falls back to the number for the host drive only. In any case, the "real" number is easily extracted by tools or terminal commannds.

But he is not using those tools but Activity Monitor which means the values don't tell you everything. Also as I have pointed out before some of the numbers the tools are given just don't make sense. One produced a totally insane 8500 TBW for the drive. My comment back then on that was:

The TB*100/percentage = TBW crosscheck produces at worst 6827 (170*100/2.49) TBW for the drive. More over If 170TB written is 2% then 170TB x 50 would be 100% which is 8500 TBW,

Ok that is nuts, there is no way there are 6827 TBW, much less 8500 TBW, SSDs that are sanely priced (assuming they exist at all).

Since the formula derived from the one everyone and his brother has been using to predict how long the SSDs last per "Eliminate the impossible and whatever remands no mater how improbable is the truth" the percentage must be wrong.

Ergo smartctl must be generating untrustful numbers and is therefore useless.

 

[Fred Zed]

What version of the OS do you have installed?

Big Sur 11.2.3 I believe it’s the current version.

 

[k-hawinkler]

But he is not using those tools but Activity Monistory which means the values don't tell you everything. Also as I have pointed out before some of the numbers the tools are given just don't make sense. One produced a totally insane 8500 TBW for the drive. My comment back then on that was:

The TB*100/percentage = TBW crosscheck produces at worst 6827 (170*100/2.49) TBW for the drive. More over If 170TB written is 2% then 170TB x 50 would be 100% which is 8500 TBW,

Ok that is nuts, there is no way there are 6827 TBW, much less 8500 TBW, SSDs that are sanely priced (assuming they exist at all).

Since the formula derived from the one everyone and his brother has been using to predict how long the SSDs last per "Eliminate the impossible and whatever remands no mater how improbable is the truth" the percentage must be wrong.

Ergo smartctl must be generating untrustful numbers and is therefore useless.TB*100/percentage = TBW

This formula actual TB written*100/percentage used = actual TB written/relative fraction used = estimated lifetime TBW is just a definition for projected guaranteed usage of the SSD by the manufacturer. Nothing more and nothing less.

If we assume TB to be the actual measured value and percentage used to be a manufacturers estimate, then the lifetime TBW is just a swag = suggested wild ass guess.
Garbage in, garbage out.

However, assuming for arguments sake Toshiba’s 3,000 writes per cell as being representative that gives 3,000/4 = 750 TBW as an estimate for a 256 GB SSD.
If you consider that value to high, replace it with your own favorite number and go from there.

 

[leons]

But he is not using those tools but Activity Monistory which means the values don't tell you everything. Also as I have pointed out before some of the numbers the tools are given just don't make sense. One produced a totally insane 8500 TBW for the drive. My comment back then on that was:

The TB*100/percentage = TBW crosscheck produces at worst 6827 (170*100/2.49) TBW for the drive. More over If 170TB written is 2% then 170TB x 50 would be 100% which is 8500 TBW,

Ok that is nuts, there is no way there are 6827 TBW, much less 8500 TBW, SSDs that are sanely priced (assuming they exist at all).

Since the formula derived from the one everyone and his brother has been using to predict how long the SSDs last per "Eliminate the impossible and whatever remands no mater how improbable is the truth" the percentage must be wrong.

Ergo smartctl must be generating untrustful numbers and is therefore useless.

I am not using smartctl or any other external tool. I am extracting the numbers from the SMART data in the drive itself. While there may be one or two "outlier" numbers reported by previous posters, I choose to ignore them as possible misreads or misreporting. The vast majority of the numbers reported here and elsewhere are consistent. It is, of course, possible that the numbers that Apple is reporting in SMART are incorrect. Anything is possible, but until given clear evidence that they are incorrect, I will accept them. Or, we need to disbelieve everything that Apple reports about the machine until proven. Add to that the fact that this problem has been widely reported. Apple's lack of response AFTER the problem has been reported would make them liable for a giant class-action suit. The large bank of lawyers in Cupertino would not allow that to happen. If these reported numbers were incorrect, they would have put out a correction or statement by now.

 

[leons]

This formula actual TB written*100/percentage used = actual TB written/relative fraction used = estimated lifetime TBW is just a definition for projected guaranteed usage of the SSD by the manufacturer. Nothing more and nothing less.

If we assume TB to be the actual measured value and percentage used to be a manufacturers estimate, then the lifetime TBW is just a swag = suggested wild ass guess.
Garbage in, garbage out.

However, assuming for arguments sake Toshiba’s 3,000 writes per cell as being representative that gives 3,000/4 = 750 TBW as an estimate for a 256 GB SSD.
If you consider that value to high, replace it with your own favorite number and go from there.

See my previous post about Class Action. The responsible manufacturer, in this case, is Apple (regardless of who made the drive). Apple CHOSE to define a "percent life" value via the SMART data in the drive. If the life of the drive were found to be drastically lower for a large percentage of the purchasers, an implied warranty would be asserted. There is next to a zero chance that the value (which if anything is stated as being a conservative estimate of lifespan) is substantially lower than the way it is coded in SMART AND that Apple is ignoring a real issue that has been widely reported.