Adds a new resource option `aliases` which can be used to rename a resource. When making a breaking change to the name or type of a resource or component, the old name can be added to the list of `aliases` for a resource to ensure that existing resources will be migrated to the new name instead of being deleted and replaced with the new named resource. There are two key places this change is implemented. The first is the step generator in the engine. When computing whether there is an old version of a registered resource, we now take into account the aliases specified on the registered resource. That is, we first look up the resource by its new URN in the old state, and then by any aliases provided (in order). This can allow the resource to be matched as a (potential) update to an existing resource with a different URN. The second is the core `Resource` constructor in the JavaScript (and soon Python) SDKs. This change ensures that when a parent resource is aliased, that all children implicitly inherit corresponding aliases. It is similar to how many other resource options are "inherited" implicitly from the parent. Four specific scenarios are explicitly tested as part of this PR: 1. Renaming a resource 2. Adopting a resource into a component (as the owner of both component and consumption codebases) 3. Renaming a component instance (as the owner of the consumption codebase without changes to the component) 4. Changing the type of a component (as the owner of the component codebase without changes to the consumption codebase) 4. Combining (1) and (3) to make both changes to a resource at the same time
232 lines
9.7 KiB
Go
232 lines
9.7 KiB
Go
// Copyright 2016-2018, Pulumi Corporation.
|
|
//
|
|
// Licensed under the Apache License, Version 2.0 (the "License");
|
|
// you may not use this file except in compliance with the License.
|
|
// You may obtain a copy of the License at
|
|
//
|
|
// http://www.apache.org/licenses/LICENSE-2.0
|
|
//
|
|
// Unless required by applicable law or agreed to in writing, software
|
|
// distributed under the License is distributed on an "AS IS" BASIS,
|
|
// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
|
|
// See the License for the specific language governing permissions and
|
|
// limitations under the License.
|
|
|
|
package display
|
|
|
|
import (
|
|
"encoding/json"
|
|
"fmt"
|
|
"time"
|
|
|
|
"github.com/golang/glog"
|
|
|
|
"github.com/pulumi/pulumi/pkg/apitype"
|
|
"github.com/pulumi/pulumi/pkg/diag"
|
|
"github.com/pulumi/pulumi/pkg/diag/colors"
|
|
"github.com/pulumi/pulumi/pkg/engine"
|
|
"github.com/pulumi/pulumi/pkg/resource"
|
|
"github.com/pulumi/pulumi/pkg/resource/config"
|
|
"github.com/pulumi/pulumi/pkg/resource/deploy"
|
|
"github.com/pulumi/pulumi/pkg/resource/stack"
|
|
"github.com/pulumi/pulumi/pkg/util/contract"
|
|
)
|
|
|
|
// massagePropertyValue takes a property value and strips out the secrets annotations from it. If showSecrets is
|
|
// not true any secret values are replaced with "[secret]".
|
|
func massagePropertyValue(v resource.PropertyValue, showSecrets bool) resource.PropertyValue {
|
|
switch {
|
|
case v.IsArray():
|
|
new := make([]resource.PropertyValue, len(v.ArrayValue()))
|
|
for i, e := range v.ArrayValue() {
|
|
new[i] = massagePropertyValue(e, showSecrets)
|
|
}
|
|
return resource.NewArrayProperty(new)
|
|
case v.IsObject():
|
|
new := make(resource.PropertyMap, len(v.ObjectValue()))
|
|
for k, e := range v.ObjectValue() {
|
|
new[k] = massagePropertyValue(e, showSecrets)
|
|
}
|
|
return resource.NewObjectProperty(new)
|
|
case v.IsSecret() && showSecrets:
|
|
return massagePropertyValue(v.SecretValue().Element, showSecrets)
|
|
case v.IsSecret():
|
|
return resource.NewStringProperty("[secret]")
|
|
default:
|
|
return v
|
|
}
|
|
}
|
|
|
|
// MassageSecrets takes a property map and returns a new map by transforming each value with massagePropertyValue
|
|
// This allows us to serialize the resulting map using our existing serialization logic we use for deployments, to
|
|
// produce sane output for stackOutputs. If we did not do this, SecretValues would be serialized as objects
|
|
// with the signature key and value.
|
|
func MassageSecrets(m resource.PropertyMap, showSecrets bool) resource.PropertyMap {
|
|
new := make(resource.PropertyMap, len(m))
|
|
for k, e := range m {
|
|
new[k] = massagePropertyValue(e, showSecrets)
|
|
}
|
|
return new
|
|
}
|
|
|
|
// stateForJSONOutput prepares some resource's state for JSON output. This includes filtering the output based
|
|
// on the supplied options, in addition to massaging secret fields.
|
|
func stateForJSONOutput(s *resource.State, opts Options) *resource.State {
|
|
var inputs resource.PropertyMap
|
|
var outputs resource.PropertyMap
|
|
if !isRootURN(s.URN) || !opts.SuppressOutputs {
|
|
// For now, replace any secret properties as the string [secret] and then serialize what we have.
|
|
inputs = MassageSecrets(s.Inputs, false)
|
|
outputs = MassageSecrets(s.Outputs, false)
|
|
} else {
|
|
// If we're suppressing outputs, don't show the root stack properties.
|
|
inputs = resource.PropertyMap{}
|
|
outputs = resource.PropertyMap{}
|
|
}
|
|
|
|
return resource.NewState(s.Type, s.URN, s.Custom, s.Delete, s.ID, inputs,
|
|
outputs, s.Parent, s.Protect, s.External, s.Dependencies, s.InitErrors, s.Provider,
|
|
s.PropertyDependencies, s.PendingReplacement, s.AdditionalSecretOutputs, s.Aliases)
|
|
}
|
|
|
|
// ShowJSONEvents renders engine events from a preview into a well-formed JSON document. Note that this does not
|
|
// emit events incrementally so that it can guarantee anything emitted to stdout is well-formed. This means that,
|
|
// if used interactively, the experience will lead to potentially very long pauses. If run in CI, it is up to the
|
|
// end user to ensure that output is periodically printed to prevent tools from thinking preview has hung.
|
|
func ShowJSONEvents(op string, action apitype.UpdateKind, events <-chan engine.Event, done chan<- bool, opts Options) {
|
|
// Ensure we close the done channel before exiting.
|
|
defer func() { close(done) }()
|
|
|
|
// Now loop and accumulate our digest until the event stream is closed, or we hit a cancellation.
|
|
var digest previewDigest
|
|
for e := range events {
|
|
// In the event of cancelation, break out of the loop immediately.
|
|
if e.Type == engine.CancelEvent {
|
|
break
|
|
}
|
|
|
|
// For all other events, use the payload to build up the JSON digest we'll emit later.
|
|
switch e.Type {
|
|
// Events ocurring early:
|
|
case engine.PreludeEvent:
|
|
// Capture the config map from the prelude. Note that all secrets will remain blinded for safety.
|
|
digest.Config = e.Payload.(engine.PreludeEventPayload).Config
|
|
|
|
// Events throughout the execution:
|
|
case engine.DiagEvent:
|
|
// Skip any ephemeral or debug messages, and elide all colorization.
|
|
p := e.Payload.(engine.DiagEventPayload)
|
|
if !p.Ephemeral && p.Severity != diag.Debug {
|
|
digest.Diagnostics = append(digest.Diagnostics, previewDiagnostic{
|
|
URN: p.URN,
|
|
Message: colors.Never.Colorize(p.Prefix + p.Message),
|
|
Severity: p.Severity,
|
|
})
|
|
}
|
|
case engine.StdoutColorEvent:
|
|
// Append stdout events as informational messages, and elide all colorization.
|
|
p := e.Payload.(engine.StdoutEventPayload)
|
|
digest.Diagnostics = append(digest.Diagnostics, previewDiagnostic{
|
|
Message: colors.Never.Colorize(p.Message),
|
|
Severity: diag.Info,
|
|
})
|
|
case engine.ResourcePreEvent:
|
|
// Create the detailed metadata for this step and the initial state of its resource. Later,
|
|
// if new outputs arrive, we'll search for and swap in those new values.
|
|
if m := e.Payload.(engine.ResourcePreEventPayload).Metadata; shouldShow(m, opts) || isRootStack(m) {
|
|
step := &previewStep{
|
|
Op: m.Op,
|
|
URN: m.URN,
|
|
Provider: m.Provider,
|
|
DiffReasons: m.Diffs,
|
|
ReplaceReasons: m.Keys,
|
|
}
|
|
|
|
if m.Old != nil {
|
|
oldState := stateForJSONOutput(m.Old.State, opts)
|
|
res, err := stack.SerializeResource(oldState, config.NewPanicCrypter())
|
|
if err == nil {
|
|
step.OldState = &res
|
|
} else {
|
|
glog.V(7).Infof("not adding old state as there was an error serialzing: %s", err)
|
|
}
|
|
}
|
|
if m.New != nil {
|
|
newState := stateForJSONOutput(m.New.State, opts)
|
|
res, err := stack.SerializeResource(newState, config.NewPanicCrypter())
|
|
if err == nil {
|
|
step.NewState = &res
|
|
} else {
|
|
glog.V(7).Infof("not adding new state as there was an error serialzing: %s", err)
|
|
}
|
|
}
|
|
|
|
digest.Steps = append(digest.Steps, step)
|
|
}
|
|
case engine.ResourceOutputsEvent, engine.ResourceOperationFailed:
|
|
// Because we are only JSON serializing previews, we don't need to worry about outputs
|
|
// resolving or operations failing. In the future, if we serialize actual deployments, we will
|
|
// need to come up with a scheme for matching the failure to the associated step.
|
|
|
|
// Events ocurring late:
|
|
case engine.SummaryEvent:
|
|
// At the end of the preview, a summary event indicates the final conclusions.
|
|
p := e.Payload.(engine.SummaryEventPayload)
|
|
digest.Duration = p.Duration
|
|
digest.ChangeSummary = p.ResourceChanges
|
|
digest.MaybeCorrupt = p.MaybeCorrupt
|
|
default:
|
|
contract.Failf("unknown event type '%s'", e.Type)
|
|
}
|
|
}
|
|
|
|
// Finally, go ahead and render the JSON to stdout.
|
|
out, err := json.MarshalIndent(&digest, "", " ")
|
|
contract.Assertf(err == nil, "unexpected JSON error: %v", err)
|
|
fmt.Println(string(out))
|
|
}
|
|
|
|
// previewDigest is a JSON-serializable overview of a preview operation.
|
|
type previewDigest struct {
|
|
// Config contains a map of configuration keys/values used during the preview. Any secrets will be blinded.
|
|
Config map[string]string `json:"config,omitempty"`
|
|
|
|
// Steps contains a detailed list of all resource step operations.
|
|
Steps []*previewStep `json:"steps,omitempty"`
|
|
// Diagnostics contains a record of all warnings/errors that took place during the preview. Note that
|
|
// ephemeral and debug messages are omitted from this list, as they are meant for display purposes only.
|
|
Diagnostics []previewDiagnostic `json:"diagnostics,omitempty"`
|
|
|
|
// Duration records the amount of time it took to perform the preview.
|
|
Duration time.Duration `json:"duration,omitempty"`
|
|
// ChangeSummary contains a map of count per operation (create, update, etc).
|
|
ChangeSummary engine.ResourceChanges `json:"changeSummary,omitempty"`
|
|
// MaybeCorrupt indicates whether one or more resources may be corrupt.
|
|
MaybeCorrupt bool `json:"maybeCorrupt,omitempty"`
|
|
}
|
|
|
|
// previewStep is a detailed overview of a step the engine intends to take.
|
|
type previewStep struct {
|
|
// Op is the kind of operation being performed.
|
|
Op deploy.StepOp `json:"op"`
|
|
// URN is the resource being affected by this operation.
|
|
URN resource.URN `json:"urn"`
|
|
// Provider is the provider that will perform this step.
|
|
Provider string `json:"provider,omitempty"`
|
|
// OldState is the old state for this resource, if appropriate given the operation type.
|
|
OldState *apitype.ResourceV3 `json:"oldState,omitempty"`
|
|
// NewState is the new state for this resource, if appropriate given the operation type.
|
|
NewState *apitype.ResourceV3 `json:"newState,omitempty"`
|
|
// DiffReasons is a list of keys that are causing a diff (for updating steps only).
|
|
DiffReasons []resource.PropertyKey `json:"diffReasons,omitempty"`
|
|
// ReplaceReasons is a list of keys that are causing replacement (for replacement steps only).
|
|
ReplaceReasons []resource.PropertyKey `json:"replaceReasons,omitempty"`
|
|
}
|
|
|
|
// previewDiagnostic is a warning or error emitted during the execution of the preview.
|
|
type previewDiagnostic struct {
|
|
URN resource.URN `json:"urn,omitempty"`
|
|
Prefix string `json:"prefix,omitempty"`
|
|
Message string `json:"message,omitempty"`
|
|
Severity diag.Severity `json:"severity,omitempty"`
|
|
}
|