this derivation will be built: /nix/store/4f0qfivapza4s3hp85rcw9009arkl26n-treefmt-check.drv error: Cannot build '/nix/store/4f0qfivapza4s3hp85rcw9009arkl26n-treefmt-check.drv'. Reason: builder failed with exit code 1. Output paths: /nix/store/dai6iagwakm74zv4xvrnx0d3x7ih6l7r-treefmt-check Last 25 log lines: > M content/posts/next-generation-distributed-nix-builds/index.md > diff --git a/content/posts/next-generation-distributed-nix-builds/index.md b/content/posts/next-generation-distributed-nix-builds/index.md > index 701d7ea..f540555 100644 > --- a/content/posts/next-generation-distributed-nix-builds/index.md > +++ b/content/posts/next-generation-distributed-nix-builds/index.md > @@ -41,15 +41,15 @@ ssh-ng://customer@mac02.numtide.com aarch64-darwin,x86_64-darwin /run/secrets/ss > ``` > > Scaling this to a group becomes a bit cumbersome because now everyone has to > -keep this list of machines up-to-date and also each ssh key needs to be > -deployed in the list of keys for every builder machine. > +keep this list of machines up-to-date and also each ssh key needs to be deployed > +in the list of keys for every builder machine. > > To solve this problem we introduce a `Farm mode` in nix-grpc-store: > > ![farm mode: clients, envoy, nodes and the cache](farm.svg) > > -One builder gets elected to become a build scheduler and every other worker > -will connect to this scheduler to receive build jobs. Finished builds will be > +One builder gets elected to become a build scheduler and every other worker will > +connect to this scheduler to receive build jobs. Finished builds will be > uploaded to [niks3](https://github.com/Mic92/niks3/), which is a binary cache > service that can upload to S3 compatible storage such as Garage or SeaweedFS. > For full logs, run: nix log /nix/store/4f0qfivapza4s3hp85rcw9009arkl26n-treefmt-check.drv