Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

iam: separate backoffs, add jitter, and increase for conflicts #10786

Closed
wants to merge 1 commit into from
Closed

iam: separate backoffs, add jitter, and increase for conflicts #10786

Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
30 changes: 20 additions & 10 deletions google/iam.go
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -5,6 +5,7 @@ import (
"encoding/json"
"fmt"
"log"
"math/rand"
"reflect"
"sort"
"strings"
Expand Down Expand Up @@ -80,13 +81,20 @@ func iamPolicyReadModifyWrite(updater ResourceIamUpdater, modify iamPolicyModify
mutexKV.Lock(mutexKey)
defer mutexKV.Unlock(mutexKey)

backoff := time.Second
// Used for introducing jitter in backoffs
rand.Seed(time.Now().UTC().UnixNano())

readBackoff := time.Second
conflictBackoff := time.Second
serviceAccountReadBackoff := time.Second
for {
log.Printf("[DEBUG]: Retrieving policy for %s\n", updater.DescribeResource())
p, err := updater.GetResourceIamPolicy()
if isGoogleApiErrorWithCode(err, 429) {
log.Printf("[DEBUG] 429 while attempting to read policy for %s, waiting %v before attempting again", updater.DescribeResource(), backoff)
time.Sleep(backoff)
readBackoffWithJitter := readBackoff + time.Duration(rand.Intn(1000))*time.Millisecond
log.Printf("[DEBUG] 429 while attempting to read policy for %s, waiting %v before attempting again", updater.DescribeResource(), readBackoffWithJitter)
time.Sleep(readBackoffWithJitter)
readBackoff = readBackoff * 2
continue
} else if err != nil {
return err
Expand Down Expand Up @@ -141,10 +149,11 @@ func iamPolicyReadModifyWrite(updater ResourceIamUpdater, modify iamPolicyModify
break
}
if isConflictError(err) {
log.Printf("[DEBUG]: Concurrent policy changes, restarting read-modify-write after %s\n", backoff)
time.Sleep(backoff)
backoff = backoff * 2
if backoff > 30*time.Second {
conflictBackoffWithJitter := conflictBackoff + time.Duration(rand.Intn(1000))*time.Millisecond
log.Printf("[DEBUG]: Concurrent policy changes, restarting read-modify-write after %v\n", conflictBackoffWithJitter)
time.Sleep(conflictBackoffWithJitter)
conflictBackoff = conflictBackoff * 2
if conflictBackoff > 5*time.Minute {
return errwrap.Wrapf(fmt.Sprintf("Error applying IAM policy to %s: Too many conflicts. Latest error: {{err}}", updater.DescribeResource()), err)
}
continue
Expand All @@ -160,9 +169,10 @@ func iamPolicyReadModifyWrite(updater ResourceIamUpdater, modify iamPolicyModify
if rerr != nil {
if p.Etag != currentPolicy.Etag {
// not matching indicates that there is a new state to attempt to apply
log.Printf("current and old etag did not match for %s, retrying", updater.DescribeResource())
time.Sleep(backoff)
backoff = backoff * 2
serviceAccountBackoffWithJitter := serviceAccountReadBackoff + time.Duration(rand.Intn(1000))*time.Millisecond
log.Printf("current and old etag did not match for %s, retrying after %v", updater.DescribeResource(), serviceAccountBackoffWithJitter)
time.Sleep(serviceAccountBackoffWithJitter)
serviceAccountReadBackoff = serviceAccountReadBackoff * 2
continue
}

Expand Down