bcantrill ("Bryan Cantrill") wrote:
Seven years ago -- almost to the day! -- I wrote my rather unvarnished thoughts on unikernels and their fitness for production:
http://dtrace.org/blogs/bmc/2016/01/22/unikernels-are-unfit-for-production/
I have long wanted to circle back to this topic, because our experience with both Hubris and Helios at @oxidecomputer have served to broaden perspectives. Join @ahl and me tomorrow on @oxidecomputer as we revisit unikernels:
https://discord.gg/7V9hdZzy?event=1066962259483570226
Recorded as always, and available wherever you get podcasts: