
Replying to @dominikkoch @cramforce
happy to connect in our shared slack channel (just joined it!)
Replying to @RhysSullivan @LukeberryPi
If you find a missing feature, we will ship it lickity split
Replying to @RhysSullivan @LukeberryPi
would love to know if Vercel Daemons work for you
Let’s link up and I can get you enrolled in this: vercel.com/changelog/flat-ra…
Replying to @thekitze @dansanmedina
😶 I'll take a look! Thanks for the quick reply. If there's anything I can ever do for you in the future, let me know!
Replying to @thekitze @dansanmedina
would love to know what stopped you from using vercel's $0 plan
curious how much magic people want for .NET/Java/Scala/Ruby/etc on Vercel
buildpack-style auto-detect? Dockerfiles you can C+P and edit? or no magic, just a generic Dockerfile option like today?
is there a better way?
Replying to @_pi0_
awesome! Managed-runtime Functions (i.e. Node.js) that run on our new microVM architecture will also benefit from this streaming soon™️
Replying to @rauchg
not today but that's a really good idea. it might not be a large lift to collect "number of bytes read before app listened on PORT"
and a metric for total bytes read over full session, so we can suggest how to trim image size
my favorite engineering detail about Dockerfile on Vercel is that boots are streaming, so startup time does not increase with container size. it only matters how many bytes your app reads before it listens on PORT
You can now run any Dockerfile on Vercel.
# 𝙳𝚘𝚌𝚔𝚎𝚛𝚏𝚒𝚕𝚎.𝚟𝚎𝚛𝚌𝚎𝚕
𝙵𝚁𝙾𝙼 𝚐𝚘𝚕𝚊𝚗𝚐:𝟷.𝟸𝟺
𝙲𝙾𝙿𝚈 . .
𝚁𝚄𝙽 𝚐𝚘 𝚋𝚞𝚒𝚕𝚍 -𝚘 /𝚜𝚎𝚛𝚟𝚎𝚛 .
𝙲𝙼𝙳 ["/𝚜𝚎𝚛𝚟𝚎𝚛"]
vercel.com/blog/dockerfile-o…
Docker is back! It’s been incredible to work alongside the team bringing this back to life.
xitter.cf/vercel/status/20719510…
Replying to @jineshshah36
What kind of utilization are you achieving across your GKE fleet (actual CPU, not allocated CPU)?
Replying to @Southclaws
Is the container image unique per customer, or is it the same image spun up in a dedicated env per customer?


