feat: add dynamic timeout retry with file-size-based parameters #468
There are no files selected for viewing
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| Original file line number | Diff line number | Diff line change |
|---|---|---|
|
|
@@ -17,14 +17,20 @@ | |
| package backend | ||
|
|
||
| import ( | ||
| legacymodelspec "github.com/dragonflyoss/model-spec/specs-go/v1" | ||
| modelspec "github.com/modelpack/model-spec/specs-go/v1" | ||
| ) | ||
|
|
||
| // getAnnotationFilepath returns the filepath stored on a descriptor's | ||
| // annotations, preferring the modelpack key and falling back to the legacy | ||
| // dragonflyoss key so older artifacts remain readable. Returns empty string | ||
| // when neither key is present. | ||
| func getAnnotationFilepath(annotations map[string]string) string { | ||
| if annotations == nil { | ||
| return "" | ||
| } | ||
| if path := annotations[modelspec.AnnotationFilepath]; path != "" { | ||
| return path | ||
| } | ||
| return annotations[legacymodelspec.AnnotationFilepath] | ||
| } | ||
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| Original file line number | Diff line number | Diff line change |
|---|---|---|
|
|
@@ -23,7 +23,6 @@ import ( | |
| "os" | ||
| "path/filepath" | ||
|
|
||
| modelspec "github.com/modelpack/model-spec/specs-go/v1" | ||
| ocispec "github.com/opencontainers/image-spec/specs-go/v1" | ||
| "github.com/sirupsen/logrus" | ||
|
|
@@ -35,6 +34,7 @@ import ( | |
| "github.com/modelpack/modctl/pkg/backend/processor" | ||
| "github.com/modelpack/modctl/pkg/config" | ||
| "github.com/modelpack/modctl/pkg/modelfile" | ||
| "github.com/modelpack/modctl/pkg/retrypolicy" | ||
| "github.com/modelpack/modctl/pkg/source" | ||
| ) | ||
|
|
||
|
|
@@ -44,7 +44,11 @@ const ( | |
| ) | ||
|
|
||
| // Build builds the user materials into the model artifact which follows the Model Spec. | ||
| func (b *backend) Build( | ||
| ctx context.Context, | ||
| modelfilePath, workDir, target string, | ||
| cfg *config.Build, | ||
| ) error { | ||
| logrus.Infof("build: building artifact %s", target) | ||
| // parse the repo name and tag name from target. | ||
| ref, err := ParseReference(target) | ||
|
|
@@ -123,8 +127,8 @@ func (b *backend) Build(ctx context.Context, modelfilePath, workDir, target stri | |
|
|
||
| var configDesc ocispec.Descriptor | ||
| // Build the model config. | ||
| if err := retrypolicy.Do(ctx, func(rctx context.Context) error { | ||
| configDesc, err = builder.BuildConfig(rctx, config, hooks.NewHooks( | ||
| hooks.WithOnStart(func(name string, size int64, reader io.Reader) io.Reader { | ||
| return pb.Add(internalpb.NormalizePrompt("Building config"), name, size, reader) | ||
| }), | ||
|
|
@@ -136,13 +140,16 @@ func (b *backend) Build(ctx context.Context, modelfilePath, workDir, target stri | |
| }), | ||
| )) | ||
| return err | ||
| }, retrypolicy.DoOpts{ | ||
| FileSize: 0, // config is small | ||
| FileName: "config", | ||
| }); err != nil { | ||
| return fmt.Errorf("failed to build model config: %w", err) | ||
| } | ||
|
|
||
| // Build the model manifest. | ||
| if err := retrypolicy.Do(ctx, func(rctx context.Context) error { | ||
| _, err = builder.BuildManifest(rctx, layers, configDesc, manifestAnnotation(modelfile), hooks.NewHooks( | ||
| hooks.WithOnStart(func(name string, size int64, reader io.Reader) io.Reader { | ||
| return pb.Add(internalpb.NormalizePrompt("Building manifest"), name, size, reader) | ||
| }), | ||
|
|
@@ -154,23 +161,32 @@ func (b *backend) Build(ctx context.Context, modelfilePath, workDir, target stri | |
| }), | ||
| )) | ||
| return err | ||
| }, retrypolicy.DoOpts{ | ||
| FileSize: 0, // manifest is small | ||
| FileName: "manifest", | ||
| }); err != nil { | ||
| return fmt.Errorf("failed to build model manifest: %w", err) | ||
| } | ||
|
|
||
| logrus.Infof("build: built artifact %s", target) | ||
| return nil | ||
| } | ||
|
|
||
| func (b *backend) getProcessors( | ||
| modelfile modelfile.Modelfile, | ||
| cfg *config.Build, | ||
| ) []processor.Processor { | ||
| processors := []processor.Processor{} | ||
|
|
||
| if configs := modelfile.GetConfigs(); len(configs) > 0 { | ||
| mediaType := modelspec.MediaTypeModelWeightConfig | ||
| if cfg.Raw { | ||
| mediaType = modelspec.MediaTypeModelWeightConfigRaw | ||
| } | ||
| processors = append( | ||
| processors, | ||
| processor.NewModelConfigProcessor(b.store, mediaType, configs, ""), | ||
| ) | ||
| } | ||
|
|
||
| if models := modelfile.GetModels(); len(models) > 0 { | ||
|
|
@@ -201,10 +217,23 @@ func (b *backend) getProcessors(modelfile modelfile.Modelfile, cfg *config.Build | |
| } | ||
|
|
||
| // process walks the user work directory and process the identified files. | ||
| func (b *backend) process( | ||
| ctx context.Context, | ||
| builder build.Builder, | ||
| workDir string, | ||
| pb *internalpb.ProgressBar, | ||
| cfg *config.Build, | ||
| processors ...processor.Processor, | ||
| ) ([]ocispec.Descriptor, error) { | ||
| descriptors := []ocispec.Descriptor{} | ||
| for _, p := range processors { | ||
| descs, err := p.Process( | ||
| ctx, | ||
| builder, | ||
| workDir, | ||
| processor.WithConcurrency(cfg.Concurrency), | ||
| processor.WithProgressTracker(pb), | ||
| ) | ||
| if err != nil { | ||
| return nil, err | ||
| } | ||
|
|
||
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| Original file line number | Diff line number | Diff line change |
|---|---|---|
|
|
@@ -19,18 +19,20 @@ package backend | |
| import ( | ||
| "context" | ||
| "encoding/json" | ||
| "errors" | ||
| "fmt" | ||
| "sync" | ||
| "time" | ||
|
|
||
| "github.com/bmatcuk/doublestar/v4" | ||
| ocispec "github.com/opencontainers/image-spec/specs-go/v1" | ||
| "github.com/sirupsen/logrus" | ||
| "golang.org/x/sync/errgroup" | ||
|
|
||
| internalpb "github.com/modelpack/modctl/internal/pb" | ||
| "github.com/modelpack/modctl/pkg/backend/remote" | ||
| "github.com/modelpack/modctl/pkg/config" | ||
| "github.com/modelpack/modctl/pkg/retrypolicy" | ||
| ) | ||
|
|
||
| // Fetch fetches partial files to the output. | ||
|
|
@@ -74,10 +76,7 @@ func (b *backend) Fetch(ctx context.Context, target string, cfg *config.Fetch) e | |
| for _, layer := range manifest.Layers { | ||
| for _, pattern := range cfg.Patterns { | ||
| if anno := layer.Annotations; anno != nil { | ||
| path := getAnnotationFilepath(anno) | ||
| // Use doublestar.PathMatch for pattern matching to support ** recursive matching | ||
| // PathMatch uses the system's native path separator (like filepath.Match) while | ||
| // also supporting recursive patterns like **/*.json | ||
|
|
@@ -101,9 +100,12 @@ func (b *backend) Fetch(ctx context.Context, target string, cfg *config.Fetch) e | |
| pb.Start() | ||
| defer pb.Stop() | ||
|
|
||
| g := new(errgroup.Group) | ||
| g.SetLimit(cfg.Concurrency) | ||
|
|
||
| var mu sync.Mutex | ||
| var errs []error | ||
|
|
||
| logrus.Infof("fetch: fetching %d matched layers", len(layers)) | ||
| for _, layer := range layers { | ||
| g.Go(func() error { | ||
|
|
@@ -113,17 +115,37 @@ func (b *backend) Fetch(ctx context.Context, target string, cfg *config.Fetch) e | |
| default: | ||
| } | ||
|
|
||
| annoFilepath := getAnnotationFilepath(layer.Annotations) | ||
|
|
||
| logrus.Debugf("fetch: processing layer %s", layer.Digest) | ||
| if err := retrypolicy.Do(ctx, func(rctx context.Context) error { | ||
| return pullAndExtractFromRemote(rctx, pb, internalpb.NormalizePrompt("Fetching blob"), client, cfg.Output, layer) | ||
| }, retrypolicy.DoOpts{ | ||
| FileSize: layer.Size, | ||
| FileName: annoFilepath, | ||
| OnRetry: func(attempt uint, reason string, backoff time.Duration) { | ||
| if bar := pb.Get(layer.Digest.String()); bar != nil { | ||
| bar.SetRefill(bar.Current()) | ||
| bar.SetCurrent(0) | ||
| bar.EwmaSetCurrent(0, time.Second) | ||
| } | ||
| }, | ||
|
Comment on lines
+126
to
+132
There was a problem hiding this comment. Use the OnRetry: func(attempt uint, reason string, backoff time.Duration) {
pb.Placeholder(layer.Digest.String(), internalpb.NormalizePrompt("Fetching blob"), layer.Size)
}, |
||
| }); err != nil { | ||
| mu.Lock() | ||
| errs = append(errs, err) | ||
| mu.Unlock() | ||
| } else { | ||
| logrus.Debugf("fetch: successfully processed layer %s", layer.Digest) | ||
| } | ||
| return nil | ||
| }) | ||
| } | ||
|
|
||
| _ = g.Wait() | ||
| if ctx.Err() != nil { | ||
| return fmt.Errorf("fetch cancelled: %w", ctx.Err()) | ||
| } | ||
| if err := errors.Join(errs...); err != nil { | ||
| return err | ||
| } | ||
|
|
||
|
|
||
Oops, something went wrong.
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
The
Placeholdermethod should encapsulate the full reset logic used during retries, including setting the refill value and resetting the EWMA speed calculation. This ensures consistency across different transfer paths and avoids manual bar manipulation in the backend packages.