3 ms·

otoh: very few workload are static. our world isn’t (circadian rhythm is a thing!) and this drives most of our systems not to be such.

• kortilla 4y ago
  Static on a daily rolling average. You’ll be hemorrhaging money if you think moving to the cloud will save you money because you have a daily trough. The only places I’ve seen it work are: - massive seasonal demand with negligible usage during off-season (think TurboTax if it weren’t part of Intuit) - brand new product development where you don’t even know what you need yet or you are going through exponential growth that you can’t get ahead of. (This stage should only last like 6 months though)
  • snotrockets 4y ago
    You are doing measuring wrong (look at distributions, not a number), and moving to the cloud wrong (rearchitect with elasticity and serverless in mind, don’t just put it on instances). I was hoping those lessons would be a bit more common by now.
    • kortilla 4y ago
      You don’t understand how the vast majority of compute workloads pan out for regular businesses if you think “elasticity” has any relevance at all.
    • ReflectedImage 4y ago
      Well at a company I worked at, on the surface it looked like we only needed to be running compute 9 am - 8 pm. But nowadays, it runs full load 24/7 on the cloud. Turns out the after hours batch jobs are just as demanding. shrugs