4 ms·

Apple resolved M1 Mac SSD wear reporting issue in macOS 11.4

21 points by quyleanh 5y ago

SXX 5y ago
What an awful journalism. They cite Marcan's tweets, but title this "news" as: > Apple resolved M1 Mac SSD wear reporting issue in macOS 11.4 Seriously Marcan himself said many times reporting was always correct. E.g check the topic they cited: https://twitter.com/marcan42/status/1396374313591140357 https://twitter.com/marcan42/status/1396374313591140357
vlovich123 5y ago
After all that hubbub, turns out it was just a bug in reporting the values and not anything wrong with the SSD? That’s surprising. Will be interesting to see a technical post mortem if one is ever written
- tonyedgecombe 5y ago
  >That’s surprising. It wasn't to me. The anti-Apple stuff can become quite rabid, even on HN.
  - SXX 5y ago
    How is it anti-Apple stuff when both smartctl and Apple's own Activity Monitor reporting the same values? And this article cite Marcan saying it's "reporting issue" while Marcan himself said multiple times it's not.
    - tonyedgecombe 5y ago
      You are kind of proving my point.
      - SXX 5y ago
        What is anti-Apple in calling excessive swapping for what it is? I own Macbook on M1 and I had 4% wearoff in two months of use (25TBW on 256GB model) when problem got attention so I could get rid of x86 apps which caused worse swapping.
        vlovich123 5y ago
        Except isn’t the details trickling out here suggests there wasn’t excessive swapping and the value you’re relying on to conclude that is erroneous? Did you somehow measure swapping differently?
        SXX 5y ago
        Check Hector Martin (marcan42) replies on Twitter that AppleInsider quotes. He verified that it's not just erroneous reporting.
    - codyb 5y ago
      Wouldn’t two programs reporting the same value just imply some common os or hardware level function being used by both? It can certainly bolster a belief to have corroborating evidence though.
      - SXX 5y ago
        It's not just two programs reporting the same value. It's SSD controller reporting TBW amout that grow exactly as fast as specific kernel process writes grow in Activity monitor. smartctl just reporting information that SSD controller returns from it's own SMART data. So this data even stored independently from an operation system. Activity Monitor on other hand actually watch all system IO of currently running processes and keep record of how much each process writes / reads from / to disk while being monitored.
sixbrx 5y ago
Not sure the quotes support the article's contention that is only a "reporting" issue. Hector Martin's quote calls it the "SSD thrashing issue" not a reporting issue, and says diffing the kernel source will be interesting to see what the issue was - would only really be "interesting" IMO if it were thrashing, not a simple reporting problem.
- GeekyBear 5y ago
  >Not sure the quotes support the article's contention that is only a "reporting" issue. I agree. From his responses to questions in the related Twitter thread, he is of the opinion that while the SSD thrashing will stop going forward, this wasn't simply a reporting bug. >The update isn't going to magically undo the existing wear. https://twitter.com/marcan42/status/1400845974197530627 https://twitter.com/marcan42/status/1400845974197530627

21 points by quyleanh 5y ago

SXX 5y ago

What an awful journalism. They cite Marcan's tweets, but title this "news" as: > Apple resolved M1 Mac SSD wear reporting issue in macOS 11.4 Seriously Marcan himself said many times reporting was always correct. E.g check the topic they cited: https://twitter.com/marcan42/status/1396374313591140357 https://twitter.com/marcan42/status/1396374313591140357

vlovich123 5y ago

After all that hubbub, turns out it was just a bug in reporting the values and not anything wrong with the SSD? That’s surprising. Will be interesting to see a technical post mortem if one is ever written

tonyedgecombe 5y ago
>That’s surprising. It wasn't to me. The anti-Apple stuff can become quite rabid, even on HN.
- SXX 5y ago
  How is it anti-Apple stuff when both smartctl and Apple's own Activity Monitor reporting the same values? And this article cite Marcan saying it's "reporting issue" while Marcan himself said multiple times it's not.
  - tonyedgecombe 5y ago
    You are kind of proving my point.
    - SXX 5y ago
      What is anti-Apple in calling excessive swapping for what it is? I own Macbook on M1 and I had 4% wearoff in two months of use (25TBW on 256GB model) when problem got attention so I could get rid of x86 apps which caused worse swapping.
      - vlovich123 5y ago
        Except isn’t the details trickling out here suggests there wasn’t excessive swapping and the value you’re relying on to conclude that is erroneous? Did you somehow measure swapping differently?
        SXX 5y ago
        Check Hector Martin (marcan42) replies on Twitter that AppleInsider quotes. He verified that it's not just erroneous reporting.
  - codyb 5y ago
    Wouldn’t two programs reporting the same value just imply some common os or hardware level function being used by both? It can certainly bolster a belief to have corroborating evidence though.
    - SXX 5y ago
      It's not just two programs reporting the same value. It's SSD controller reporting TBW amout that grow exactly as fast as specific kernel process writes grow in Activity monitor. smartctl just reporting information that SSD controller returns from it's own SMART data. So this data even stored independently from an operation system. Activity Monitor on other hand actually watch all system IO of currently running processes and keep record of how much each process writes / reads from / to disk while being monitored.

sixbrx 5y ago

Not sure the quotes support the article's contention that is only a "reporting" issue. Hector Martin's quote calls it the "SSD thrashing issue" not a reporting issue, and says diffing the kernel source will be interesting to see what the issue was - would only really be "interesting" IMO if it were thrashing, not a simple reporting problem.

GeekyBear 5y ago
>Not sure the quotes support the article's contention that is only a "reporting" issue. I agree. From his responses to questions in the related Twitter thread, he is of the opinion that while the SSD thrashing will stop going forward, this wasn't simply a reporting bug. >The update isn't going to magically undo the existing wear. https://twitter.com/marcan42/status/1400845974197530627 https://twitter.com/marcan42/status/1400845974197530627