I never use more than a very few browser tabs at the same time (Safari and Chrome) yet I have one of the worst cases of this issue so far with 149 TB written in just 67 days.
Since I disabled Spotlight completely the average data written per day went from 2.225 TB to "just" 483 GB, so there seems to be a big difference there. Still high, but there is indeed the difference.
As for the sleep theory, I am not convinced yet that the problem is with the sleep itself because it's not like hibernate so it's not supposed to write much to disk when activated. It should just keep the content it already has in memory and the virtual memory (swap) as is.
Having said that, I put the computer to sleep at 0:44 yesterday and woke it today at 13:10, so 12 hours and 26 minutes. In this period of time, a whopping 230 GB of data have been written to the SSD (I am monitoring and tracking all the numbers constantly in a spreadsheet to calculate/estimate stuff).
So it's a ridiculous amount of data written during sleep and it shouldn't happen. But I don't think it's because of the sleep itself, but because of "something" that is triggered during sleep and writes to disk.
Besides Spotlight, I am now wondering if TimeMachine's local snapshots also play a role in all of this. AFAIK, TimeMachine does a partial wake up from sleep (without activating the screens) in order to perform its backups. Even when using an external drive for backups, by default TimeMachine keeps local snapshots on the main drive too, so it has to write something there. But the only process I am aware of related to TimeMachine is backupd, which for me has written only 9.85 GB so far since I rebooted yesterday. I am not sure of whether something in kernel_task is also involved in TimeMachine backups. Does anyone know?
According to my calculations, 10% of lifetime used at 149 TB for my 256 GB equals to a 100% lifetime of 1.488 PB.
If the computer continued to write an average 2.225 TB / day as it has done in 66 days, the 100% lifetime would be reached on around October 27, 2022.
If it continues to write around 483 GB / day as it's doing since disabling Spotlight completely (estimation based on almost 15 hours so far so this might change), the 100% lifetime would be reached on around October 5, 2028, which is a massive difference.
Can anyone confirm that the 100% lifetime for a 256 GB drive would be 1.488 PB? The way I have calculated it is simple: 291990055 units have been written so far at 10% of used lifetime for a total of 149 TB. So each unit is around 510-511 KB.
So 100% lifetime would be around 2919900550 units * 511 KB = 1.49 PB.
Am I calculating this wrongly?
Thanks.
Is there some reason not to use the mdutil command instead?
sudo mdutil -i off /Volumes/<Data>
I haven't needed to do this but it seems like disabling a system daemon is overkill when there is a standard way to do the same thing.
With the mdutil thing only the mds process was still running for me, while it has disappeared after disabling the daemon.