ICP.Ninja Problem

Hi
for some time I’ve been having an issue deploying canisters from ICP.Ninja. I tested it on different machines and I still get the same message:

[ Building and deploying code onchain: Calm: Failed to reserve ephemeral canisters: error in response: status code 422 Unprocessable Entity: Failed to batch create canisters: Failed to create canister due to 'Rate limit exceeded: Rate limit exceeded for action create_ephemeral_canister' ]

Additionally, I noticed that a canister I created in ICP.Ninja some time ago is still running on the subnet (unless I’m mistaken, it should have been removed after 30 minutes).

Jessie is gone and so is Vivienne. Elias is team caffeine now, not even sure who to tag on this one :thinking:
@Henn91 also got some errors couple of days ago.

Ninja hasn’t been receiving a lot of attention lately.
I’ll take a look at it.

Since you guys are here, @KryL, @ZackDS and @Henn91 could you answer some informal questions:

  • What are you currently working on?
  • How often do you use ninja?
  • What you like about ninja?
  • What you don’t like about ninja?
  • Have you tried the local developer experience?
  • How is the local developer experience relative to ninja? When you would pick one over the other?

I’d appreciate your feedback - if you prefer to message me in private feel free to do so.

Should be back up as of a few hours ago if you want to give it a try.

The short version is that:

  • There is a “calm” service - it’s what manages the pool of canisters and allocates them when folks are trying to run something through ninja.
  • A couple of months ago the calm service was changed to decouple from caffeine
  • The cron jobs that return used canisters to the pool were not running

Should be fixed now.

FYI - ninja is still only compatible with dfx projects. We are looking at updating all the example to work with icp-cli and updating ninja to use icp-cli too.

will edit this in the morning … reminder (still need to take it off one doc)