Build cache is specific to individual builder machines. We use sticky builds so your service consistently lands on the same machine, but occasionally a builder gets cycled out for upgrades, which means the cache is lost. This build is installing everything from scratch, and chromium in particular is a heavy package. Once this build completes, your cache will be re-established on the new builder and subsequent builds should return to normal speed.

I'm seeing the same thing with a very similar application, again which has been building successfully for weeks. This behaviour is completely unlike what I've seen up to now from deploys on Railway. I have the same application deployed across your US and EU servers. The EU servers slowed down in the late EU morning, while the US was still deploying smoothly and quickly. Now the problem is occurring on both sides of the pond. Has something changed on the Railway side?

Please open your own thread. Jumping onto someone else's thread does not allow us to provide assistance to individual users in an efficient manner.