Promtail: Fix retry/stop when erroring for out of cloudflare retention range (e.g. over 168 hours old) (#5698)

* fix: 5696 don't retry when out of range for cloudflare

CHANGELOG.md

@@ -1,4 +1,5 @@
 ## Main
+* [5696](https://github.com/grafana/loki/pull/5696) **paullryan** don't block scraping of new logs from cloudflare within promtail if an error is received from cloudflare about too early logs.
 * [5685](https://github.com/grafana/loki/pull/5625) **chaudum** Fix bug in push request parser that allowed users to send arbitrary non-string data as "log line".
 * [5707](https://github.com/grafana/loki/pull/5707) **franzwong** Promtail: Rename config name limit_config to limits_config.
 * [5626](https://github.com/grafana/loki/pull/5626) **jeschkies** Support multi-tenant select logs and samples queries.

clients/pkg/promtail/targets/cloudflare/target.go

@@ -3,6 +3,7 @@ package cloudflare
 import (
 	"context"
 	"errors"
+	"regexp"
 	"strings"
 	"sync"
 	"time"
@@ -28,6 +29,8 @@ import (
 // The minimun window size is 1 minute.
 const minDelay = time.Minute
 
+var cloudflareTooEarlyError = regexp.MustCompile(`too early: logs older than \S+ are not available`)
+
 var defaultBackoff = backoff.Config{
 	MinBackoff: 1 * time.Second,
 	MaxBackoff: 10 * time.Second,
@@ -151,7 +154,10 @@ func (t *Target) pull(ctx context.Context, start, end time.Time) error {
 
 	for backoff.Ongoing() {
 		it, err = t.client.LogpullReceived(ctx, start, end)
-		if err != nil {
+		if err != nil && cloudflareTooEarlyError.MatchString(err.Error()) {
+			level.Warn(t.logger).Log("msg", "failed iterating over logs, out of cloudflare range, not retrying", "err", err, "start", start, "end", end, "retries", backoff.NumRetries())
+			return nil
+		} else if err != nil {
 			errs.Add(err)
 			backoff.Wait()
 			continue

clients/pkg/promtail/targets/cloudflare/target_test.go

@@ -200,6 +200,58 @@ func Test_CloudflareTargetError(t *testing.T) {
 	require.Equal(t, newEnd, end.UnixNano())
 }
 
+func Test_CloudflareTargetError168h(t *testing.T) {
+	var (
+		w      = log.NewSyncWriter(os.Stderr)
+		logger = log.NewLogfmtLogger(w)
+		cfg    = &scrapeconfig.CloudflareConfig{
+			APIToken:  "foo",
+			ZoneID:    "bar",
+			Labels:    model.LabelSet{"job": "cloudflare"},
+			PullRange: model.Duration(time.Minute),
+		}
+		end      = time.Unix(0, time.Hour.Nanoseconds())
+		client   = fake.New(func() {})
+		cfClient = newFakeCloudflareClient()
+	)
+	ps, err := positions.New(logger, positions.Config{
+		SyncPeriod:    10 * time.Second,
+		PositionsFile: t.TempDir() + "/positions.yml",
+	})
+	// retries as fast as possible.
+	defaultBackoff.MinBackoff = 0
+	defaultBackoff.MaxBackoff = 0
+
+	// set our end time to be the last time we have a position
+	ps.Put(positions.CursorKey(cfg.ZoneID), end.UnixNano())
+	require.NoError(t, err)
+
+	// setup errors for all retries
+	cfClient.On("LogpullReceived", mock.Anything, mock.Anything, mock.Anything).Return(nil, errors.New("HTTP status 400: bad query: error parsing time: invalid time range: too early: logs older than 168h0m0s are not available"))
+	// replace the client.
+	getClient = func(apiKey, zoneID string, fields []string) (Client, error) {
+		return cfClient, nil
+	}
+
+	ta, err := NewTarget(NewMetrics(prometheus.NewRegistry()), logger, client, ps, cfg)
+	require.NoError(t, err)
+	require.True(t, ta.Ready())
+
+	// wait for the target to be stopped.
+	require.Eventually(t, func() bool {
+		return cfClient.CallCount() >= 5
+	}, 5*time.Second, 100*time.Millisecond)
+
+	require.Len(t, client.Received(), 0)
+	require.GreaterOrEqual(t, cfClient.CallCount(), 5)
+	ta.Stop()
+	ps.Stop()
+
+	// Make sure we move on from the save the last position.
+	newEnd, _ := ps.Get(positions.CursorKey(cfg.ZoneID))
+	require.Greater(t, newEnd, end.UnixNano())
+}
+
 func Test_validateConfig(t *testing.T) {
 	tests := []struct {
 		in      *scrapeconfig.CloudflareConfig