Skip to content

Latest commit

 

History

History
164 lines (135 loc) · 11.3 KB

README.md

File metadata and controls

164 lines (135 loc) · 11.3 KB
Raw

grade

grade ingests Go benchmark data into InfluxDB so that you can track performance over time.

Installation

To download and install the grade executable into your $GOPATH/bin:

go get github.com/influxdata/grade/cmd/grade

Usage

Initial database configuration

The data from Go benchmarks tends to be very time-sparse (up to perhaps dozens of commits per day), so we recommend creating your database with an infinite retention and a large shard duration. Issue this command to your InfluxDB instance:

CREATE DATABASE benchmarks WITH DURATION INF SHARD DURATION 90d

Running the command

Although you can pipe the output of go test directly into grade, for now we recommend placing the output of go test in a file first so that if something goes wrong, you don't have to wait again to run all the benchmarks.

For example, to run all the benchmarks in your current Go project:

go test -run=^$ -bench=. -benchmem ./... > bench.txt

Then, assuming you are in the directory of your Go project and git has checked out the same commit corresponding with the tests that have run, this is the bare set of options to load the benchmark results into InfluxDB via grade:

grade \
  -hardwareid="my dev machine" \
  -goversion="$(go version | cut -d' ' -f3-)" \
  -revision="$(git log -1 --format=%H)" \
  -timestamp="$(git log -1 --format=%ct)" \
  -branch="$(git rev-parse --abbrev-ref HEAD)" \
  < bench.txt

Notes on this style of invocation:

  • -influxurl is not provided but defaults to http://localhost:8086. Basic auth credentials can be embedded in the URL if needed. HTTPS is supported; supply -insecure if you need to skip SSL verification. If you set it to an empty string, grade will print line protocol to stdout.
  • -database is not provided but defaults to benchmarks.
  • -measurement is not provided but defaults to go.
  • The hardware ID is a string that you specify to identify the hardware on which the benchmarks were run.
  • The Go version subcommand will produce a string like go1.6.2 darwin/amd64, but you can use any string you'd like.
  • The revision subcommand is the full SHA of the commit, but feel free to use a git tag name or any other string.
  • The timestamp is a Unix epoch timestamp in seconds. The above subcommand produces the Unix timestamp for the committer of the most recent commit. This assumes that the commits whose benchmarks are being run, all are ascending in time; git does not enforce that commits' timestamps are ascending, so if this assumption is broken, your data may look strange when you visualize it.
  • The branch subcommand is the name of the current branch. The -branch flag is optional.

Schema

For each benchmark result from a run of go test -bench:

  • Tags:
    • goversion is the same string as passed in to the -goversion flag.
    • hwid is the same string as passed in to the -hardwareid flag.
    • name is the name of the benchmark function, stripped of the Benchmark prefix.
    • pkg is the name of Go package containing the benchmark, e.g. github.com/influxdata/influxdb/services/httpd.
    • procs is the number of CPUs used to run the benchmark. This is a tag because you are more likely to group by procs rather than chart them over time.
    • branch is the same string as passed in to the -branch flag. Since the -branch flag is optional and can be omited, the tag will be present only if the flag is set.
  • Fields:
    • alloced_bytes_per_op is the allocated bytes per iteration of the benchmark.
    • allocs_per_op is how many allocations occurred per iteration of the benchmark.
    • mb_per_s is how many megabytes processed per second when running the benchmark.
    • n is the number of iterations in the benchmark.
    • ns_per_op is the number of wall nanoseconds taken per iteration of the benchmark.
    • revision is the git revision specified in the -revision flag. This was chosen to be a field so that the information is quickly available but not at the cost of a growing series cardinality per benchmark run.

Sample

For a benchmark like this:

PASS
BenchmarkMarshal-2                  	  500000	      2901 ns/op	     560 B/op	      13 allocs/op
BenchmarkParsePointNoTags-2         	 2000000	       733 ns/op	  31.36 MB/s	     208 B/op	       4 allocs/op
BenchmarkParsePointWithPrecisionN-2 	 2000000	       627 ns/op	  36.68 MB/s	     208 B/op	       4 allocs/op
BenchmarkParsePointWithPrecisionU-2 	 2000000	       636 ns/op	  36.15 MB/s	     208 B/op	       4 allocs/op
BenchmarkParsePointsTagsSorted2-2   	 2000000	       947 ns/op	  53.85 MB/s	     240 B/op	       4 allocs/op
BenchmarkParsePointsTagsSorted5-2   	 1000000	      1189 ns/op	  69.75 MB/s	     272 B/op	       4 allocs/op
BenchmarkParsePointsTagsSorted10-2  	 1000000	      1624 ns/op	  88.05 MB/s	     320 B/op	       4 allocs/op
BenchmarkParsePointsTagsUnSorted2-2 	 1000000	      1167 ns/op	  43.69 MB/s	     272 B/op	       5 allocs/op
BenchmarkParsePointsTagsUnSorted5-2 	 1000000	      1627 ns/op	  50.99 MB/s	     336 B/op	       5 allocs/op
BenchmarkParsePointsTagsUnSorted10-2	  500000	      2733 ns/op	  52.32 MB/s	     448 B/op	       5 allocs/op
BenchmarkParseKey-2                 	 1000000	      2361 ns/op	    1030 B/op	      24 allocs/op
ok  	github.com/influxdata/influxdb/models	19.809s

Which is passed to grade like this:

grade \
  -influxurl '' \
  -goversion "$(go version | cut -d' ' -f3-)" \
  -hardwareid c4.large \
  -revision v1.0.2 \
  -timestamp "$(cd $GOPATH/src/github.com/influxdata/influxdb && git log v1.0.2 -1 --format=%ct)" \
  -branch="$(git rev-parse --abbrev-ref HEAD)" \
  < models-1.0.2.txt

You will see output like:

go,branch=master,goversion=go1.6.2\ linux/amd64,hwid=c4.large,name=Marshal,pkg=github.com/influxdata/influxdb/models,procs=2 alloced_bytes_per_op=560i,allocs_per_op=13i,n=500000i,ns_per_op=2901,revision="v1.0.2" 1475695157000000000
go,branch=master,goversion=go1.6.2\ linux/amd64,hwid=c4.large,name=ParsePointNoTags,pkg=github.com/influxdata/influxdb/models,procs=2 alloced_bytes_per_op=208i,allocs_per_op=4i,mb_per_s=31.36,n=2000000i,ns_per_op=733,revision="v1.0.2" 1475695157000000000
go,branch=master,goversion=go1.6.2\ linux/amd64,hwid=c4.large,name=ParsePointWithPrecisionN,pkg=github.com/influxdata/influxdb/models,procs=2 alloced_bytes_per_op=208i,allocs_per_op=4i,mb_per_s=36.68,n=2000000i,ns_per_op=627,revision="v1.0.2" 1475695157000000000
go,branch=master,goversion=go1.6.2\ linux/amd64,hwid=c4.large,name=ParsePointWithPrecisionU,pkg=github.com/influxdata/influxdb/models,procs=2 alloced_bytes_per_op=208i,allocs_per_op=4i,mb_per_s=36.15,n=2000000i,ns_per_op=636,revision="v1.0.2" 1475695157000000000
go,branch=master,goversion=go1.6.2\ linux/amd64,hwid=c4.large,name=ParsePointsTagsSorted2,pkg=github.com/influxdata/influxdb/models,procs=2 alloced_bytes_per_op=240i,allocs_per_op=4i,mb_per_s=53.85,n=2000000i,ns_per_op=947,revision="v1.0.2" 1475695157000000000
go,branch=master,goversion=go1.6.2\ linux/amd64,hwid=c4.large,name=ParsePointsTagsSorted5,pkg=github.com/influxdata/influxdb/models,procs=2 alloced_bytes_per_op=272i,allocs_per_op=4i,mb_per_s=69.75,n=1000000i,ns_per_op=1189,revision="v1.0.2" 1475695157000000000
go,branch=master,goversion=go1.6.2\ linux/amd64,hwid=c4.large,name=ParsePointsTagsSorted10,pkg=github.com/influxdata/influxdb/models,procs=2 alloced_bytes_per_op=320i,allocs_per_op=4i,mb_per_s=88.05,n=1000000i,ns_per_op=1624,revision="v1.0.2" 1475695157000000000
go,branch=master,goversion=go1.6.2\ linux/amd64,hwid=c4.large,name=ParsePointsTagsUnSorted2,pkg=github.com/influxdata/influxdb/models,procs=2 alloced_bytes_per_op=272i,allocs_per_op=5i,mb_per_s=43.69,n=1000000i,ns_per_op=1167,revision="v1.0.2" 1475695157000000000
go,branch=master,goversion=go1.6.2\ linux/amd64,hwid=c4.large,name=ParsePointsTagsUnSorted5,pkg=github.com/influxdata/influxdb/models,procs=2 alloced_bytes_per_op=336i,allocs_per_op=5i,mb_per_s=50.99,n=1000000i,ns_per_op=1627,revision="v1.0.2" 1475695157000000000
go,branch=master,goversion=go1.6.2\ linux/amd64,hwid=c4.large,name=ParsePointsTagsUnSorted10,pkg=github.com/influxdata/influxdb/models,procs=2 alloced_bytes_per_op=448i,allocs_per_op=5i,mb_per_s=52.32,n=500000i,ns_per_op=2733,revision="v1.0.2" 1475695157000000000
go,branch=master,goversion=go1.6.2\ linux/amd64,hwid=c4.large,name=ParseKey,pkg=github.com/influxdata/influxdb/models,procs=2 alloced_bytes_per_op=1030i,allocs_per_op=24i,n=1000000i,ns_per_op=2361,revision="v1.0.2" 1475695157000000000

Or without the -branch flag:

grade \
  -influxurl '' \
  -goversion "$(go version | cut -d' ' -f3-)" \
  -hardwareid c4.large \
  -revision v1.0.2 \
  -timestamp "$(cd $GOPATH/src/github.com/influxdata/influxdb && git log v1.0.2 -1 --format=%ct)" \
  < models-1.0.2.txt

You will see output like:

go,goversion=go1.6.2\ linux/amd64,hwid=c4.large,name=Marshal,pkg=github.com/influxdata/influxdb/models,procs=2 alloced_bytes_per_op=560i,allocs_per_op=13i,n=500000i,ns_per_op=2901,revision="v1.0.2" 1475695157000000000
go,goversion=go1.6.2\ linux/amd64,hwid=c4.large,name=ParsePointNoTags,pkg=github.com/influxdata/influxdb/models,procs=2 alloced_bytes_per_op=208i,allocs_per_op=4i,mb_per_s=31.36,n=2000000i,ns_per_op=733,revision="v1.0.2" 1475695157000000000
go,goversion=go1.6.2\ linux/amd64,hwid=c4.large,name=ParsePointWithPrecisionN,pkg=github.com/influxdata/influxdb/models,procs=2 alloced_bytes_per_op=208i,allocs_per_op=4i,mb_per_s=36.68,n=2000000i,ns_per_op=627,revision="v1.0.2" 1475695157000000000
go,goversion=go1.6.2\ linux/amd64,hwid=c4.large,name=ParsePointWithPrecisionU,pkg=github.com/influxdata/influxdb/models,procs=2 alloced_bytes_per_op=208i,allocs_per_op=4i,mb_per_s=36.15,n=2000000i,ns_per_op=636,revision="v1.0.2" 1475695157000000000
go,goversion=go1.6.2\ linux/amd64,hwid=c4.large,name=ParsePointsTagsSorted2,pkg=github.com/influxdata/influxdb/models,procs=2 alloced_bytes_per_op=240i,allocs_per_op=4i,mb_per_s=53.85,n=2000000i,ns_per_op=947,revision="v1.0.2" 1475695157000000000
go,goversion=go1.6.2\ linux/amd64,hwid=c4.large,name=ParsePointsTagsSorted5,pkg=github.com/influxdata/influxdb/models,procs=2 alloced_bytes_per_op=272i,allocs_per_op=4i,mb_per_s=69.75,n=1000000i,ns_per_op=1189,revision="v1.0.2" 1475695157000000000
go,goversion=go1.6.2\ linux/amd64,hwid=c4.large,name=ParsePointsTagsSorted10,pkg=github.com/influxdata/influxdb/models,procs=2 alloced_bytes_per_op=320i,allocs_per_op=4i,mb_per_s=88.05,n=1000000i,ns_per_op=1624,revision="v1.0.2" 1475695157000000000
go,goversion=go1.6.2\ linux/amd64,hwid=c4.large,name=ParsePointsTagsUnSorted2,pkg=github.com/influxdata/influxdb/models,procs=2 alloced_bytes_per_op=272i,allocs_per_op=5i,mb_per_s=43.69,n=1000000i,ns_per_op=1167,revision="v1.0.2" 1475695157000000000
go,goversion=go1.6.2\ linux/amd64,hwid=c4.large,name=ParsePointsTagsUnSorted5,pkg=github.com/influxdata/influxdb/models,procs=2 alloced_bytes_per_op=336i,allocs_per_op=5i,mb_per_s=50.99,n=1000000i,ns_per_op=1627,revision="v1.0.2" 1475695157000000000
go,goversion=go1.6.2\ linux/amd64,hwid=c4.large,name=ParsePointsTagsUnSorted10,pkg=github.com/influxdata/influxdb/models,procs=2 alloced_bytes_per_op=448i,allocs_per_op=5i,mb_per_s=52.32,n=500000i,ns_per_op=2733,revision="v1.0.2" 1475695157000000000
go,goversion=go1.6.2\ linux/amd64,hwid=c4.large,name=ParseKey,pkg=github.com/influxdata/influxdb/models,procs=2 alloced_bytes_per_op=1030i,allocs_per_op=24i,n=1000000i,ns_per_op=2361,revision="v1.0.2" 1475695157000000000