TUN-3902: Add jitter to backoffhandler

Jitter is important to avoid every cloudflared in the world trying to
reconnect at t=1, 2, 4, etc. That could overwhelm the backend. But
if each cloudflared randomly waits for up to 2, then up to 4, then up
to 8 etc, then the retries get spread out evenly across time.

On average, wait times should be the same (e.g. instead of waiting for
exactly 1 second, cloudflared will wait betweeen 0 and 2 seconds).

This is the "Full Jitter" algorithm from https://aws.amazon.com/blogs/architecture/exponential-backoff-and-jitter/
This commit is contained in:
Adam Chalmers
2021-02-10 10:42:09 -06:00
parent 3004703074
commit a278753bbf
7 changed files with 49 additions and 38 deletions

View File

@@ -8,6 +8,7 @@ import (
"time"
"github.com/stretchr/testify/assert"
"github.com/stretchr/testify/require"
tunnelpogs "github.com/cloudflare/cloudflared/tunnelrpc/pogs"
)
@@ -28,13 +29,14 @@ func TestRefreshAuthBackoff(t *testing.T) {
// authentication failures should consume the backoff
for i := uint(0); i < backoff.MaxRetries; i++ {
retryChan, err := rcm.RefreshAuth(context.Background(), backoff, auth)
assert.NoError(t, err)
assert.NotNil(t, retryChan)
assert.Equal(t, (1<<i)*time.Second, wait)
require.NoError(t, err)
require.NotNil(t, retryChan)
require.Greater(t, wait.Seconds(), 0.0)
require.Less(t, wait.Seconds(), float64((1<<(i+1))*time.Second))
}
retryChan, err := rcm.RefreshAuth(context.Background(), backoff, auth)
assert.Error(t, err)
assert.Nil(t, retryChan)
require.Error(t, err)
require.Nil(t, retryChan)
// now we actually make contact with the remote server
_, _ = rcm.RefreshAuth(context.Background(), backoff, func(ctx context.Context, n int) (tunnelpogs.AuthOutcome, error) {
@@ -47,8 +49,8 @@ func TestRefreshAuthBackoff(t *testing.T) {
expectedGracePeriod := time.Duration(time.Second * 2 << backoff.MaxRetries)
return time.Now().Add(expectedGracePeriod * 2)
}
_, ok := backoff.GetBackoffDuration(context.Background())
assert.True(t, ok)
_, ok := backoff.GetMaxBackoffDuration(context.Background())
require.True(t, ok)
}
func TestRefreshAuthSuccess(t *testing.T) {