Clojurians Log v2

Maybe someone can help me here, I have a CircleCi setup with two docker containers, one running my own code, and another running a s3 mock. Now, when calling this s3 mock on my own machine, thats instant, but on CircleCi it takes 20 seconds. That sounds like some timeout/backoff being triggered, but how could I debug what is ‘different’ on CircleCI?

@lennart.buit are you testing with the local circleci CLI command?

No I am not, maybe worth a try, didn’t even know that was a thing

it’s especially handy for testing your config before deploying (which I do too often to test…)

yeah, I am using it to validate, but I didn’t expect that it actually could run tests for me

thats nifty!

I don’t experience the same slowness locally through the CLI, the first invocation to the S3 mock is ‘slow’ at ~6 seconds, the last is alright at 700ms. At least nowhere near the 20 seconds I am seeing on CircleCi. So somewhere I am paying a 20 second penalty on Circle

are the logs public?

no — sorry

good luck then 😉

yeah 😞

thanks for the help tho!

maybe it’s downloading/doing stuff that could be cached?

or maybe you could avoid the mock by talking to the real S3 with a test bucket?

Yeah, I probably should do the latter, but the purist in me is complaining :’). Its so strange because it actually works, its not just failing, its just succeeding very slowly

I have a repo with an S3 mock (localstack) and the calls to it don't take 20s, if that helps in any way: https://circleci.com/gh/polymeris/cljs-aws/442

yeah also using localstack, that will be a great help indeed!

This is the config https://github.com/polymeris/cljs-aws/blob/master/.circleci/config.yml#L54