1. What Is Prefix Cache In model inference scenarios, caching mechanisms are often used to improve throughput and performance. There are two common caching mechanisms: Key-Value Cache (KV Cache), which is oriented toward the internals of a single request, caching the intermediate computation results (Key and Value) of the Transformer model to avoid redundant computation.

1. About JuiceFS Caching On a host, warmed cache is placed directly on the host. In a cluster, there are two levels of cache: Worker, which provides cache shared at the cluster level Fuse, which provides cache at the current node level only 2. Warming Data with the JuiceFS Client Note that warming at the Fuse layer is only effective for the current node; if you need to warm the whole cluster, you must warm at the Worker layer.

In a non-staged build scenario, when we use a container to build, we can mount the cache directory inside the container onto the build host, run the build task, and then copy the artifacts into the runtime image to produce the application image. But with a staged build, the build image and the runtime image live in the same Dockerfile, which makes it harder to optimize the caching of third-party dependencies.

In a CICD flow, there are two main kinds of things that need to be saved: build artifacts and caches. Build artifacts are the final execution result, and caches exist to speed up the next build. This post mainly describes how to archive build artifacts and caches in Jenkins, along with a hands-on example backed by object storage.