Remote Execution
Automatically parallelize your build actions and test runs across thousands of cores.
siggi@lunchbox: ~/tensorflow
siggi@lunchbox: ~/tensorflow$ bazel build tensorflow --config=remote
INFO: Invocation ID: 5fa9285b-b01e-435b-9ad4-96543f939e6a
INFO: Streaming build results to: https://app.buildbuddy.io/invocation/5fa9285b-b01e-435b-9ad4-96543f939e6a
INFO: Analyzed target //tensorflow:tensorflow (226 packages loaded, 20226 targets configured)
INFO: Found 1 target...
[0 / 8,191] 199 actions, 159 running
INFO: Invocation ID: 5fa9285b-b01e-435b-9ad4-96543f939e6a
INFO: Streaming build results to: https://app.buildbuddy.io/invocation/5fa9285b-b01e-435b-9ad4-96543f939e6a
INFO: Analyzed target //tensorflow:tensorflow (226 packages loaded, 20226 targets configured)
INFO: Found 1 target...
[0 / 8,191] 199 actions, 159 running
Compiling src/google/protobuf/io/zero_copy_stream_impl_lite.cc; 2s remote
Compiling src/google/protobuf/inlined_string_field.cc; 2s remote
Compiling src/google/protobuf/wire_format_lite.cc; 2s remote
Compiling sc/google/protobuf/stubs/strutil.cc; 2s remote
Compiling src/google/protobuf/any_lite.cc; 2s remote
Compiling src/google/protobuf/message_lite.cc; 2s remote
GoToolchainBinaryCompile external/go_sdk_Linux/builder.a [for host]; 2s remote
Compiling src/google/protobuf/stubs/structurally_valid.cc; 2s remote
Simple configuration
Just copy a few lines to your .bazelrc, and you're off to the races.
Auto-scaling
Scale your executor cluster up and down auto-magically based on load using custom metrics.
Custom docker images
Pick your distro, install the tools you need. All you need to do is specify your own custom docker image.
Pre-emptible machines
Our executors were designed from the ground up to run on preemptible machines. This means you can take advantage of Spot VM pricing.
Remote persistent workers
Need a warm JVM? Our executors support remote persistent workers to speed up JIT compilation.
Bring your own executors
Bringing your own executors gives you all of the advantages of BuildBuddy Cloud while still using your own hardware.
Live action view
Watch remotely executing actions live, as they're happening.
Action explorer
View input files, command details, environment variables, timing information, output files, and more for each individual action.
Action timeline
Dive into the timing breakdown of individual actions to see where the time is being spent.
Mac executors
Intel and M1 executor binaries allow you to run Mac-native executors for iOS builds and more.
