Awesome
goleak
Goroutine leak detector to help avoid Goroutine leaks.
Installation
You can use go get
to get the latest version:
go get -u go.uber.org/goleak
goleak
also supports semver releases.
Note that go-leak only supports the two most recent minor versions of Go.
Quick Start
To verify that there are no unexpected goroutines running at the end of a test:
func TestA(t *testing.T) {
defer goleak.VerifyNone(t)
// test logic here.
}
Instead of checking for leaks at the end of every test, goleak
can also be run
at the end of every test package by creating a TestMain
function for your
package:
func TestMain(m *testing.M) {
goleak.VerifyTestMain(m)
}
Note
For tests that use t.Parallel, goleak
does
not know how to distinguish a leaky goroutine from tests that have not finished running.
func TestA(t *testing.T) {
tt := struct{
name string
input SomeType
expected string
}{
// ...
}
for _, t := range tt {
t.Run(t.name, func(t *testing.T) {
t.Parallel() // <- goleak gets confused here!
// ...
}
}
}
For such cases you should also defer to using goleak.VerifyTestMain
as shown above.
Determine Source of Package Leaks
When verifying leaks using TestMain
, the leak test is only run once after all tests
have been run. This is typically enough to ensure there's no goroutines leaked from
tests, but when there are leaks, it's hard to determine which test is causing them.
You can use the following bash script to determine the source of the failing test:
# Create a test binary which will be used to run each test individually
$ go test -c -o tests
# Run each test individually, printing "." for successful tests, or the test name
# for failing tests.
$ for test in $(go test -list . | grep -E "^(Test|Example)"); do ./tests -test.run "^$test\$" &>/dev/null && echo -n "." || echo -e "\n$test failed"; done
This will only print names of failing tests which can be investigated individually. E.g.,
.....
TestLeakyTest failed
.......
Stability
goleak is v1 and follows SemVer strictly.
No breaking changes will be made to exported APIs before 2.0.