Cherry-pick elastic#10177 to 6.x: common.Backoff now implements jitte…

…r instead of sleeping for a fixed amount of time (elastic#10229)

Cherry-pick of PR elastic#10177 to 6.x branch. Original message: 

This PR add a new interface called backoff.Backoff, this can be used to
generalize any backoff interaction. It move the current Backoff strategy
under an ExpBackoff type.

ExpBackoff is the same as before on every wait we just
exponentially increase the duration of the wait and sleep for that
amount.

EqualJitterBackoff uses an exponential increment of the duration but
will take half of that value as fixed sleep time and the other half
as a jitter. This will help distribute the new request when a cluster is
done instead of having all the beats trying to reconnect at once.

The Redis implementations and any clients wrapped with a backoff will
now use the EqualJitterBackoff, any other code will keep using the same
exponential strategy.

Fixes: elastic#10172

reader/journal.go

@@ -35,6 +35,7 @@ import (
 	"github.com/elastic/beats/journalbeat/config"
 	"github.com/elastic/beats/libbeat/beat"
 	"github.com/elastic/beats/libbeat/common"
+	"github.com/elastic/beats/libbeat/common/backoff"
 	"github.com/elastic/beats/libbeat/logp"
 )
 
@@ -44,7 +45,7 @@ type Reader struct {
 	config  Config
 	done    chan struct{}
 	logger  *logp.Logger
-	backoff *common.Backoff
+	backoff backoff.Backoff
 }
 
 // New creates a new journal reader and moves the FP to the configured position.
@@ -98,7 +99,7 @@ func newReader(logger *logp.Logger, done chan struct{}, c Config, journal *sdjou
 		config:  c,
 		done:    done,
 		logger:  logger,
-		backoff: common.NewBackoff(done, c.Backoff, c.MaxBackoff),
+		backoff: backoff.NewExpBackoff(done, c.Backoff, c.MaxBackoff),
 	}
 	r.seek(state.Cursor)