• JPAKx4@lemmy.sdf.org
    link
    fedilink
    arrow-up
    53
    ·
    1 year ago

    There is increased corruption as the lifespan of the cache grows, maintaining speed but reducing element quality

  • fibojoly@sh.itjust.works
    link
    fedilink
    arrow-up
    35
    ·
    edit-2
    1 year ago

    This is acceptable for commonly used, hot items. But you also need a colder storage for stuff that you ain’t gonna wear everyday!
    This young lady is just inefficient and has no care for the cost of hot cache.

    Personally, I use a top of bed hot cache, a chair warm cache and the rest in cold closet.

    • InputZero@lemmy.ml
      link
      fedilink
      arrow-up
      14
      ·
      1 year ago

      I use to run my closet asynchronously like that, but I would loose way too many small packets waiting for the cache to fill. Especially when they were mirrored. Now I exclusively use synchronous writes to the cold store. It may be slower but it’s worth it for better data integrity.

      Oh pro tip, if you use a FAT filesystem avoid horizontal striping.

  • tias
    link
    fedilink
    arrow-up
    28
    arrow-down
    1
    ·
    edit-2
    1 year ago

    This is unironically how I think. Also carrying single items to their designated place is inefficient. I wait until there’s a pile so I can process in batch and avoid latency overhead. Same for vacuum cleaning, I need to strike a thoughtful balance between accumulated dust and the time expended on vacuuming. All to maximize throughput in my life.

    • Dyskolos@lemmy.zip
      link
      fedilink
      arrow-up
      9
      ·
      edit-2
      1 year ago

      Same. You can’t imagine the amount of discussions with wifey i had about this. Life is hard 🙄

      Also it’s not chaos when i exactly know where everything is. Even if there are socks on top of it. I put them there. For a reason.

  • xmunk@sh.itjust.works
    link
    fedilink
    arrow-up
    18
    ·
    1 year ago

    Personally I assemble my clothes in a B+ tree on my desk. I’ve found that the hash lookup algorithm I was using had too many collisions.