cloudflare / lasso Goto Github PK

The three lasso Workers are currently deployed manually, when someone clones the repo and runs wrangler publish. We should use a GitHub Action instead.

first check kv
if not in kv go to github
if found in github store in kv
if not in kv or github, return error

Several issues in wrangler have surfaced due to this. I haven't been able to reproduce the failure from the production worker https://workers.cloudflare.com/get-npm-wrangler-binary/1.10.3/x86_64-apple-darwin, though others have.

I cloned lasso and wrangler publish the caching worker to https://natdavidson.com/cache-wrangler-bin. Hitting https://natdavidson.com/cache-wrangler-bin/1.11.0/x86_64-apple-darwin in the browser produces could not find cache-wrangler-bin/1.11.0/x86_64-apple-darwin/index.html in your content namespace. curling the url returns

HTTP/2 500
date: Wed, 14 Oct 2020 16:29:12 GMT
content-type: text/plain;charset=UTF-8
content-length: 92
set-cookie: __cfduid=d260f6729ff3a3658aaffbe2596a194a41602692952; expires=Fri, 13-Nov-20 16:29:12 GMT; path=/; domain=.natdavidson.com; HttpOnly; SameSite=Lax; Secure
cf-request-id: 05c98aca66000029dee0181000000001
expect-ct: max-age=604800, report-uri="https://report-uri.cloudflare.com/cdn-cgi/beacon/expect-ct"
report-to: {"endpoints":[{"url":"https:\/\/a.nel.cloudflare.com\/report?lkg-colo=16&lkg-time=1602692953"}],"group":"cf-nel","max_age":604800}
nel: {"report_to":"cf-nel","max_age":604800}
server: cloudflare
cf-ray: 5e22ad8a394a29de-IAD

wrangler tail doesn't log anything.

there's lots of shared logic, no reason for three endpoints to be three different workers