1. 20
  1. 16

    In fact, Pingora crashes are so rare we usually find unrelated issues when we do encounter one… We’ve also discovered hardware issues on a few machines, in the past ruling out rare memory bugs caused by our software even after significant debugging was nearly impossible.

    Let’s define a server software to be hardware-stable if its crashes are more likely due to hardware errors than software errors. Apparently Pingora is hardware-stable while Cloudflare’s previous system was not. Hardware-stability is a concrete goal to aim for server softwares.

    You don’t necessarily need Rust to achieve hardware-stability, although it helps. Famously, Redis crashes are more likely due to hardware errors (memory in particular) than software errors, so Redis is hardware-stable. In fact, Redis includes a built-in memory tester that runs on crashes, in order to avoid spurious bug reports, because it happens so often.

    1. 1

      Closed source means it could be all bullshit.