NAME

cabe - Command-line tool for working with cabe

SYNOPSIS

cabe

[--help|-h]
[--logging]

DESCRIPTION

cli tools for analyzing and debugging pinpoint A/B experiment tryobs using cabe

Usage:

cabe [GLOBAL OPTIONS] command [COMMAND OPTIONS] [ARGUMENTS...]

GLOBAL OPTIONS

--help, -h: show help

--logging: Turn on logging while running commands.

COMMANDS

analyze

analyze runs the analyzer process locally and pretty-prints a table of results to stdout.

--benchmark="": name of benchmark to analyze

--pinpoint-job="": ID of the pinpoint job to check

--record-to-zip="": Zip file to save replay data to

--replay-from-zip="": Zip file to replay data from

--workload="": comma separated list of names of benchmark workloads to analyze

check

prints diagnostic results, and an inferred experiment spec if none was specified

--benchmark="": name of benchmark to analyze

--pinpoint-job="": ID of the pinpoint job to check

--record-to-zip="": Zip file to save replay data to

--replay-from-zip="": Zip file to replay data from

--workload="": comma separated list of names of benchmark workloads to analyze

readcas

readcas reads perf results json data from RBE-CAS, located using the provided root-digest.

--benchmark="": name of benchmark to analyze

--cas-instance="": cas instance (default: projects/chrome-swarming/instances/default_instance)

--pinpoint-job="": ID of the pinpoint job to check

--record-to-zip="": Zip file to save replay data to

--replay-from-zip="": Zip file to replay data from

--root-digest="": CAS digest for the root node

--workload="": comma separated list of names of benchmark workloads to analyze

sandwich

sandwich executes a verification workflow based on an existing pinpoint bisect job.

--attempts="": iterations verification job will run (default: 30)

--benchmark="": name of benchmark to analyze

--dry-run: dry run for StartWorkflow; just print CreateExecutionRequest to stdout

--execution="": execution id of the workflow

--location="": location for workflow execution (default: us-central1)

--pinpoint-job="": ID of the pinpoint job to check

--project="": GAE project app name (default: chromeperf)

--record-to-zip="": Zip file to save replay data to

--replay-from-zip="": Zip file to replay data from

--workflow="": name of workflow to execute (default: sandwich-verification-workflow-prod)

--workload="": comma separated list of names of benchmark workloads to analyze

markdown

Generates markdown help for cabe.

help, h

Shows a list of commands or help for one command