summaryrefslogtreecommitdiffhomepage
path: root/benchmarks/harness/machine_producers/gcloud_producer.py
AgeCommit message (Collapse)Author
2020-02-06Add logic to run from baked images.Zach Koopmans
Change adds the following: - logic to run from "baked images". See [GVISOR_DIR]/tools/images - installers which install modified files from a workspace. This allows users to run benchmarks while modifying runsc. - removes the --preemptible tag from built GCE instances. Preemptible instances are much more likely to be preempted on startup, which manifests for the user as a failed benchmark. I don't currently have a way to detect if a VM has been preempted that will work for this change. https://cloud.google.com/compute/docs/instances/preemptible#preemption_process https://cloud.google.com/compute/docs/instances/preemptible#preemption_selection PiperOrigin-RevId: 293697949
2020-01-16Add run-gcp command.Zach Koopmans
Add command to run benchmarks on GCP backed machines using the gcloud producer. Run with: `bazel run :benchmarks -- run-gcp [BENCHMARK_NAME]` Tested with the startup benchmark. PiperOrigin-RevId: 290126444
2020-01-09Implement gcloud_producer for bm-toolsZach Koopmans
bm-tools works via "Machine" objects, which are front objects for actual virtual or physical machines. glcoud_producer produces machines on GCP using the `gcloud` tool. Included are: - GCloudProducer - the class producing machines - MockGCloudProducer - class for mocking GCloudProducer. - tests using the mock - test data as .json files. Code to generate this test data may be included in a follow up. PiperOrigin-RevId: 289005958