Skip to main content
Every Aspect Workflows deployment includes a managed remote cache: REv2-compliant, deployed inside your own VPC right next to your CI runners, and operated by Aspect engineers under our SLA.

How caching works

Bazel skips re-executing any action whose inputs, command, and environment match a prior run; it fetches the cached output instead. The --remote_cache flag points Bazel at a content-addressable store so that cache hits are shared across every developer and CI job on your team, not just within a single machine.

In your VPC, next to your runners

A remote cache is only as fast as the network between it and your builds. Hosted caches put the internet in that path: every artifact pays round-trip latency, and every byte pays egress. The Workflows cache runs inside a dedicated Virtual Private Cloud, colocated with your runners, so reads and writes stay on your network at your network’s speed. It never shares quota or storage with other Workflows customers, and your build traffic never leaves the private network boundary.

Build without the Bytes

The cache supports Bazel’s --remote_download_minimal mode: instead of downloading every artifact in the build graph, Bazel fetches only those strictly required to complete the build, typically just the leaf nodes. This dramatically reduces network transfer and improves overall build efficiency.

Developer-facing access

Build clients outside the Workflows-managed runners, such as developer laptops, a secondary CI provider, or ad-hoc jobs, can connect to the same cache, either privately (VPC peering) or over the internet via the external remote cache and execution feature. External access authenticates through your SSO provider using OIDC/OAuth, with fine-grained authorization options such as read-only access for end users.

Standard protocol, no lock-in

The cache implements the Bazel Remote Execution Protocol v2 over gRPC. Any REv2-compatible client works with it, and nothing about your build depends on the cache to keep working.

Observability

Grafana dashboards in your deployment track cache hit rates, throughput, and storage over time, so you can catch performance regressions and validate optimizations.