Files
opentf/internal/tofu/node_resource_destroy.go
Martin Atkins 0503163e28 tracing: Centralize our OpenTelemetry package imports
OpenTelemetry has various Go packages split across several Go modules that
often need to be carefully upgraded together. And in particular, we are
using the "semconv" package in conjunction with the OpenTelemetry SDK's
"resource" package in a way that requires that they both agree on which
version of the OpenTelemetry Semantic Conventions are being followed.

To help avoid "dependency hell" situations when upgrading, this centralizes
all of our direct calls into the OpenTelemetry SDK and tracing API into
packages under internal/tracing, by exposing a few thin wrapper functions
that other packages can use to access the same functionality indirectly.

We only use a relatively small subset of the OpenTelemetry library surface
area, so we don't need too many of these reexports and they should not
represent a significant additional maintenance burden.

For the semconv and resource interaction in particular this also factors
that out into a separate helper function with a unit test, so we should
notice quickly whenever they become misaligned. This complements the
end-to-end test previously added in opentofu/opentofu#3447 to give us
faster feedback about this particular problem, while the end-to-end test
has the broader scope of making sure there aren't any errors at all when
initializing OpenTelemetry tracing.

Finally, this also replaces the constants we previously had in package
traceaddrs with functions that return attribute.KeyValue values directly.
This matches the API style used by the OpenTelemetry semconv packages, and
makes the calls to these helpers from elsewhere in the system a little
more concise.

Signed-off-by: Martin Atkins <mart@degeneration.co.uk>
2025-10-30 13:27:10 -07:00

292 lines
9.9 KiB
Go

// Copyright (c) The OpenTofu Authors
// SPDX-License-Identifier: MPL-2.0
// Copyright (c) 2023 HashiCorp, Inc.
// SPDX-License-Identifier: MPL-2.0
package tofu
import (
"context"
"fmt"
"log"
"github.com/opentofu/opentofu/internal/addrs"
"github.com/opentofu/opentofu/internal/communicator/shared"
"github.com/opentofu/opentofu/internal/configs"
"github.com/opentofu/opentofu/internal/instances"
"github.com/opentofu/opentofu/internal/plans"
"github.com/opentofu/opentofu/internal/states"
"github.com/opentofu/opentofu/internal/tfdiags"
"github.com/opentofu/opentofu/internal/tracing"
"github.com/opentofu/opentofu/internal/tracing/traceattrs"
)
// NodeDestroyResourceInstance represents a resource instance that is to be
// destroyed.
type NodeDestroyResourceInstance struct {
*NodeAbstractResourceInstance
// If DeposedKey is set to anything other than states.NotDeposed then
// this node destroys a deposed object of the associated instance
// rather than its current object.
DeposedKey states.DeposedKey
}
var (
_ GraphNodeModuleInstance = (*NodeDestroyResourceInstance)(nil)
_ GraphNodeConfigResource = (*NodeDestroyResourceInstance)(nil)
_ GraphNodeResourceInstance = (*NodeDestroyResourceInstance)(nil)
_ GraphNodeDestroyer = (*NodeDestroyResourceInstance)(nil)
_ GraphNodeDestroyerCBD = (*NodeDestroyResourceInstance)(nil)
_ GraphNodeReferenceable = (*NodeDestroyResourceInstance)(nil)
_ GraphNodeReferencer = (*NodeDestroyResourceInstance)(nil)
_ GraphNodeExecutable = (*NodeDestroyResourceInstance)(nil)
_ GraphNodeProviderConsumer = (*NodeDestroyResourceInstance)(nil)
_ GraphNodeProvisionerConsumer = (*NodeDestroyResourceInstance)(nil)
)
func (n *NodeDestroyResourceInstance) Name() string {
if n.DeposedKey != states.NotDeposed {
return fmt.Sprintf("%s (destroy deposed %s)", n.ResourceInstanceAddr(), n.DeposedKey)
}
return n.ResourceInstanceAddr().String() + " (destroy)"
}
func (n *NodeDestroyResourceInstance) ProvidedBy() RequestedProvider {
switch n.Addr.Resource.Resource.Mode {
case addrs.DataResourceMode:
// indicate that this node does not require a configured provider
return RequestedProvider{}
case addrs.EphemeralResourceMode:
// Since ephemeral resources are not stored into the state or plan files,
// a change of type delete cannot be generated for it, meaning that this
// code path is not meant to be reached.
// Even though, let's ensure that ever the case, a destroy node for an
// ephemeral resource indicates correctly that for its removal there
// is no provider needed.
return RequestedProvider{}
}
return n.NodeAbstractResourceInstance.ProvidedBy()
}
// GraphNodeDestroyer
func (n *NodeDestroyResourceInstance) DestroyAddr() *addrs.AbsResourceInstance {
addr := n.ResourceInstanceAddr()
return &addr
}
// GraphNodeDestroyerCBD
func (n *NodeDestroyResourceInstance) CreateBeforeDestroy() bool {
// State takes precedence during destroy.
// If the resource was removed, there is no config to check.
// If CBD was forced from descendent, it should be saved in the state
// already.
if s := n.instanceState; s != nil {
if s.Current != nil {
return s.Current.CreateBeforeDestroy
}
}
if n.Config != nil && n.Config.Managed != nil {
return n.Config.Managed.CreateBeforeDestroy
}
return false
}
// GraphNodeDestroyerCBD
func (n *NodeDestroyResourceInstance) ModifyCreateBeforeDestroy(v bool) error {
return nil
}
// GraphNodeReferenceable, overriding NodeAbstractResource
func (n *NodeDestroyResourceInstance) ReferenceableAddrs() []addrs.Referenceable {
normalAddrs := n.NodeAbstractResourceInstance.ReferenceableAddrs()
destroyAddrs := make([]addrs.Referenceable, len(normalAddrs))
phaseType := addrs.ResourceInstancePhaseDestroy
if n.CreateBeforeDestroy() {
phaseType = addrs.ResourceInstancePhaseDestroyCBD
}
for i, normalAddr := range normalAddrs {
switch ta := normalAddr.(type) {
case addrs.Resource:
destroyAddrs[i] = ta.Phase(phaseType)
case addrs.ResourceInstance:
destroyAddrs[i] = ta.Phase(phaseType)
default:
destroyAddrs[i] = normalAddr
}
}
return destroyAddrs
}
// GraphNodeReferencer, overriding NodeAbstractResource
func (n *NodeDestroyResourceInstance) References() []*addrs.Reference {
// If we have a config, then we need to include destroy-time dependencies
if c := n.Config; c != nil && c.Managed != nil {
var result []*addrs.Reference
// We include conn info and config for destroy time provisioners
// as dependencies that we have.
for _, p := range c.Managed.Provisioners {
schema := n.ProvisionerSchemas[p.Type]
if p.When == configs.ProvisionerWhenDestroy {
if p.Connection != nil {
result = append(result, ReferencesFromConfig(p.Connection.Config, shared.ConnectionBlockSupersetSchema)...)
}
result = append(result, ReferencesFromConfig(p.Config, schema)...)
}
}
return result
}
return nil
}
// GraphNodeExecutable
func (n *NodeDestroyResourceInstance) Execute(ctx context.Context, evalCtx EvalContext, op walkOperation) (diags tfdiags.Diagnostics) {
addr := n.ResourceInstanceAddr()
ctx, span := tracing.Tracer().Start(
ctx, traceNameApplyResourceInstance,
tracing.SpanAttributes(
traceattrs.String(traceAttrResourceInstanceAddr, addr.String()),
),
)
defer span.End()
// Eval info is different depending on what kind of resource this is
switch addr.Resource.Resource.Mode {
case addrs.ManagedResourceMode:
diags = n.resolveProvider(ctx, evalCtx, false, states.NotDeposed)
if diags.HasErrors() {
tracing.SetSpanError(span, diags)
return diags
}
span.SetAttributes(
traceattrs.String(traceAttrProviderInstanceAddr, traceProviderInstanceAddr(n.ResolvedProvider.ProviderConfig, n.ResolvedProviderKey)),
)
diags = diags.Append(
n.managedResourceExecute(ctx, evalCtx),
)
case addrs.DataResourceMode:
diags = diags.Append(
n.dataResourceExecute(ctx, evalCtx),
)
case addrs.EphemeralResourceMode:
diags = diags.Append(
n.ephemeralResourceExecute(ctx, evalCtx),
)
default:
panic(fmt.Errorf("unsupported resource mode %s", n.Config.Mode))
}
tracing.SetSpanError(span, diags)
return diags
}
func (n *NodeDestroyResourceInstance) managedResourceExecute(ctx context.Context, evalCtx EvalContext) (diags tfdiags.Diagnostics) {
addr := n.ResourceInstanceAddr()
// Get our state
is := n.instanceState
if is == nil {
log.Printf("[WARN] NodeDestroyResourceInstance for %s with no state", addr)
}
// These vars are updated through pointers at various stages below.
var changeApply *plans.ResourceInstanceChange
var state *states.ResourceInstanceObject
_, providerSchema, err := getProvider(ctx, evalCtx, n.ResolvedProvider.ProviderConfig, n.ResolvedProviderKey)
diags = diags.Append(err)
if diags.HasErrors() {
return diags
}
changeApply, err = n.readDiff(evalCtx, providerSchema)
diags = diags.Append(err)
if changeApply == nil || diags.HasErrors() {
return diags
}
changeApply = reducePlan(addr.Resource, changeApply, true)
// reducePlan may have simplified our planned change
// into a NoOp if it does not require destroying.
if changeApply == nil || changeApply.Action == plans.NoOp {
return diags
}
state, readDiags := n.readResourceInstanceState(ctx, evalCtx, addr)
diags = diags.Append(readDiags)
if diags.HasErrors() {
return diags
}
// Exit early if the state object is null after reading the state
if state == nil || state.Value.IsNull() {
return diags
}
diags = diags.Append(n.preApplyHook(evalCtx, changeApply))
if diags.HasErrors() {
return diags
}
// Run destroy provisioners if not tainted
if state.Status != states.ObjectTainted {
applyProvisionersDiags := n.evalApplyProvisioners(ctx, evalCtx, state, false, configs.ProvisionerWhenDestroy)
diags = diags.Append(applyProvisionersDiags)
// keep the diags separate from the main set until we handle the cleanup
if diags.HasErrors() {
// If we have a provisioning error, then we just call
// the post-apply hook now.
diags = diags.Append(n.postApplyHook(evalCtx, state, diags.Err()))
return diags
}
}
// Managed resources need to be destroyed, while data sources
// are only removed from state.
// we pass a nil configuration to apply because we are destroying
s, d := n.apply(ctx, evalCtx, state, changeApply, nil, instances.RepetitionData{}, false)
state, diags = s, diags.Append(d)
// we don't return immediately here on error, so that the state can be
// finalized
err = n.writeResourceInstanceState(ctx, evalCtx, state, workingState)
if err != nil {
return diags.Append(err)
}
// create the err value for postApplyHook
diags = diags.Append(n.postApplyHook(evalCtx, state, diags.Err()))
diags = diags.Append(updateStateHook(evalCtx, n.Addr))
return diags
}
func (n *NodeDestroyResourceInstance) dataResourceExecute(_ context.Context, evalCtx EvalContext) (diags tfdiags.Diagnostics) {
log.Printf("[TRACE] NodeDestroyResourceInstance: removing state object for %s", n.Addr)
evalCtx.State().SetResourceInstanceCurrent(n.Addr, nil, n.ResolvedProvider.ProviderConfig, n.ResolvedProviderKey)
return diags.Append(updateStateHook(evalCtx, n.Addr))
}
// ephemeralResourceExecute for NodeDestroyResourceInstance is only here to return an error.
// An ephemeral resource, by definition, cannot be destroyed. If the execution path is reaching this part, it means that
// there is an issue somewhere else, most probably in the planning phase since the generation of NodeDestroyResourceInstance
// is strictly related to the changes from the plan.
func (n *NodeDestroyResourceInstance) ephemeralResourceExecute(_ context.Context, _ EvalContext) (diags tfdiags.Diagnostics) {
log.Printf("[TRACE] NodeDestroyResourceInstance: called for ephemeral resource %s", n.Addr)
return diags.Append(tfdiags.Sourceless(
tfdiags.Error,
"Destroy invoked for an ephemeral resource",
fmt.Sprintf("A destroy operation has been invoked for the ephemeral resource %q. This is an OpenTofu error. Please report this.", n.Addr),
))
}