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

detect Retry-After during async “does resource exist?” flow #2688

Merged
merged 1 commit into from
Oct 5, 2022
Merged

detect Retry-After during async “does resource exist?” flow #2688

Show file tree
Hide file tree
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
44 changes: 38 additions & 6 deletions azure/services/async/async.go
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -18,8 +18,11 @@ package async

import (
"context"
"net/http"
"strconv"
"time"

"github.com/Azure/go-autorest/autorest"
azureautorest "github.com/Azure/go-autorest/autorest/azure"
"github.com/pkg/errors"
infrav1 "sigs.k8s.io/cluster-api-provider-azure/api/v1beta1"
Expand Down Expand Up @@ -76,7 +79,7 @@ func processOngoingOperation(ctx context.Context, scope FutureScope, client Futu

// Operation is still in progress, update conditions and requeue.
log.V(2).Info("long running operation is still ongoing", "service", serviceName, "resource", resourceName)
return nil, azure.WithTransientError(azure.NewOperationNotDoneError(future), retryAfter(sdkFuture))
return nil, azure.WithTransientError(azure.NewOperationNotDoneError(future), getRequeueAfterFromFuture(sdkFuture))
}
if err != nil {
log.V(2).Error(err, "error checking long running operation status after it finished")
Expand Down Expand Up @@ -111,7 +114,8 @@ func (s *Service) CreateResource(ctx context.Context, spec azure.ResourceSpecGet
// Get the resource if it already exists, and use it to construct the desired resource parameters.
var existingResource interface{}
if existing, err := s.Creator.Get(ctx, spec); err != nil && !azure.ResourceNotFound(err) {
return nil, errors.Wrapf(err, "failed to get existing resource %s/%s (service: %s)", rgName, resourceName, serviceName)
errWrapped := errors.Wrapf(err, "failed to get existing resource %s/%s (service: %s)", rgName, resourceName, serviceName)
return nil, azure.WithTransientError(errWrapped, getRetryAfterFromError(err))
} else if err == nil {
existingResource = existing
log.V(2).Info("successfully got existing resource", "service", serviceName, "resource", resourceName, "resourceGroup", rgName)
Expand All @@ -136,7 +140,7 @@ func (s *Service) CreateResource(ctx context.Context, spec azure.ResourceSpecGet
return nil, errors.Wrapf(err, "failed to create resource %s/%s (service: %s)", rgName, resourceName, serviceName)
}
s.Scope.SetLongRunningOperationState(future)
return nil, azure.WithTransientError(azure.NewOperationNotDoneError(future), retryAfter(sdkFuture))
return nil, azure.WithTransientError(azure.NewOperationNotDoneError(future), getRequeueAfterFromFuture(sdkFuture))
} else if err != nil {
return nil, errors.Wrapf(err, "failed to create resource %s/%s (service: %s)", rgName, resourceName, serviceName)
}
Expand Down Expand Up @@ -170,7 +174,7 @@ func (s *Service) DeleteResource(ctx context.Context, spec azure.ResourceSpecGet
return errors.Wrapf(err, "failed to delete resource %s/%s (service: %s)", rgName, resourceName, serviceName)
}
s.Scope.SetLongRunningOperationState(future)
return azure.WithTransientError(azure.NewOperationNotDoneError(future), retryAfter(sdkFuture))
return azure.WithTransientError(azure.NewOperationNotDoneError(future), getRequeueAfterFromFuture(sdkFuture))
} else if err != nil {
if azure.ResourceNotFound(err) {
// already deleted
Expand All @@ -183,12 +187,40 @@ func (s *Service) DeleteResource(ctx context.Context, spec azure.ResourceSpecGet
return nil
}

// retryAfter returns the max between the `RETRY-AFTER` header and the default requeue time.
// getRequeueAfterFromFuture returns the max between the `RETRY-AFTER` header and the default requeue time.
// This ensures we respect the retry-after header if it is set and avoid retrying too often during an API throttling event.
func retryAfter(sdkFuture azureautorest.FutureAPI) time.Duration {
func getRequeueAfterFromFuture(sdkFuture azureautorest.FutureAPI) time.Duration {
retryAfter, _ := sdkFuture.GetPollingDelay()
if retryAfter < reconciler.DefaultReconcilerRequeue {
retryAfter = reconciler.DefaultReconcilerRequeue
}
return retryAfter
}

// getRetryAfterFromError returns the time.Duration from the http.Response in the autorest.DetailedError.
// If there is no Response object, or if there is no meaningful Retry-After header data, we return a default.
func getRetryAfterFromError(err error) time.Duration {
// In case we aren't able to introspect Retry-After from the error type, we'll return this default
ret := reconciler.DefaultReconcilerRequeue
var detailedError autorest.DetailedError
// if we have a strongly typed autorest.DetailedError then we can introspect the HTTP response data
if errors.As(err, &detailedError) {
if detailedError.Response != nil {
// If we have Retry-After HTTP header data for any reason, prefer it
if retryAfter := detailedError.Response.Header.Get("Retry-After"); retryAfter != "" {
// This handles the case where Retry-After data is in the form of units of seconds
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

are we expecting both scenarios to be possible (units of seconds and absolute time)? Same question with autorest.DetailedError vs not ? Are there any SDK contracts for those error responses / headers that we can follow? Seems strange that we have to handle multiple possibilities, it's like we're guessing what the reponse might look like instead of expecting a specific format/unit/type.

Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

are we expecting both scenarios to be possible (units of seconds and absolute time)

As far as I can tell from research, the spec is overloaded to expect both value type flavors, so I think it's the best practice to deal with both types wherever we parse Retry-After HTTP header data.

Same question with autorest.DetailedError vs not

The autorest.DetailedError error implementation is one that we definitely know about from the usage above in the async flow, based on the specific SDK API we're currently re-using. In that sense you could say that this helper function is sort of tightly coupled to the particular implementation of capz at this point in time (initially this foo was inline in the CreateResource func but I split it out for code maintenance reasons). So tl;dr we only have access to Retry-After data in this particular case via the err response from the underlying service Get implementations, and we only know how to get at it (as of right now) if the err is of "type" autorest.DetailedError. Hope that makes sense!

if rai, err := strconv.Atoi(retryAfter); err == nil {
ret = time.Duration(rai) * time.Second
// This handles the case where Retry-After data is in the form of absolute time
} else if t, err := time.Parse(time.RFC1123, retryAfter); err == nil {
ret = time.Until(t)
}
// If we didn't find Retry-After HTTP header data but the response type is 429,
// we'll have to come up with our sane default.
} else if detailedError.Response.StatusCode == http.StatusTooManyRequests {
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

same here, is that a real scenario or are we doing that just in case?

Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

This is definitely a real scenario, as a autorest.DetailedError error has a strongly typed HTTP response object in it, which gives us access to the HTTP response status code.

The larger idea here is that not all API responses are well behaving, and if we get an HTTP 429 without Retry-After data we can at least trust that the HTTP 429 indicates we're sending too many requests, which is (IMO) sufficient justification for telling the controller to requeue the next request a little later in time than it would otherwise.

ret = reconciler.DefaultHTTP429RetryAfter
}
}
}
return ret
}
85 changes: 85 additions & 0 deletions azure/services/async/async_test.go
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -21,6 +21,7 @@ import (
"errors"
"net/http"
"testing"
"time"

"github.com/Azure/azure-sdk-for-go/services/resources/mgmt/2019-05-01/resources"
"github.com/Azure/go-autorest/autorest"
Expand All @@ -31,6 +32,7 @@ import (
"sigs.k8s.io/cluster-api-provider-azure/azure/mock_azure"
"sigs.k8s.io/cluster-api-provider-azure/azure/services/async/mock_async"
gomockinternal "sigs.k8s.io/cluster-api-provider-azure/internal/test/matchers/gomock"
"sigs.k8s.io/cluster-api-provider-azure/util/reconciler"
)

var (
Expand Down Expand Up @@ -434,3 +436,86 @@ func TestDeleteResource(t *testing.T) {
})
}
}

func TestGetRetryAfterFromError(t *testing.T) {
cases := []struct {
name string
input error
expected time.Duration
expectedRangeTolerance time.Duration
}{
{
name: "Retry-After header data present in the form of units of seconds",
input: autorest.DetailedError{
Response: &http.Response{
Header: http.Header{
"Retry-After": []string{"2"},
},
},
},
expected: 2 * time.Second,
},
{
name: "Retry-After header data present in the form of units of absolute time",
input: autorest.DetailedError{
Response: &http.Response{
Header: http.Header{
"Retry-After": []string{time.Now().Add(1 * time.Hour).Format(time.RFC1123)},
},
},
},
expected: 1 * time.Hour,
expectedRangeTolerance: 5 * time.Second,
},
{
name: "Retry-After header data not present",
input: autorest.DetailedError{
Response: &http.Response{
Header: http.Header{
"foo": []string{"bar"},
},
},
},
expected: reconciler.DefaultReconcilerRequeue,
},
{
name: "Retry-After header data not present in HTTP 429",
input: autorest.DetailedError{
Response: &http.Response{
StatusCode: http.StatusTooManyRequests,
Header: http.Header{
"foo": []string{"bar"},
},
},
},
expected: reconciler.DefaultHTTP429RetryAfter,
},
{
name: "nil http.Response",
input: autorest.DetailedError{
Response: nil,
},
expected: reconciler.DefaultReconcilerRequeue,
},
{
name: "error type is not autorest.DetailedError",
input: errors.New("error"),
expected: reconciler.DefaultReconcilerRequeue,
},
}

for _, c := range cases {
c := c
t.Run(c.name, func(t *testing.T) {
t.Parallel()
g := NewWithT(t)
ret := getRetryAfterFromError(c.input)
if c.expectedRangeTolerance > 0 {
g.Expect(ret).To(BeNumerically("<", c.expected))
g.Expect(ret + c.expectedRangeTolerance).To(BeNumerically(">", c.expected))
} else {
g.Expect(ret).To(Equal(c.expected))
}
})
}
}
2 changes: 2 additions & 0 deletions util/reconciler/defaults.go
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -31,6 +31,8 @@ const (
DefaultAzureCallTimeout = 2 * time.Second
// DefaultReconcilerRequeue is the default value for the reconcile retry.
DefaultReconcilerRequeue = 15 * time.Second
// DefaultHTTP429RetryAfter is a default backoff wait time when we get a HTTP 429 response with no Retry-After data.
DefaultHTTP429RetryAfter = 1 * time.Minute
)

// DefaultedLoopTimeout will default the timeout if it is zero-valued.
Expand Down