mirrors/ollama
1
0
Fork 0
mirror of https://github.com/ollama/ollama.git synced 2025-05-11 10:26:53 +02:00
Code Issues Projects Releases Packages Wiki Activity Actions

server/internal/registry: take over pulls from server package (#9485)

Browse source 
This commit replaces the old pull implementation in the server package
with the new, faster, more robust pull implementation in the registry
package.

The new endpoint, and now the remove endpoint too, are behind the
feature gate "client2" enabled only by setting the OLLAMA_EXPERIMENT
environment variable include "client2".

Currently, the progress indication is wired to perform the same as the
previous implementation to avoid making changes to the CLI, and because
the status reports happen at the start of the download, and the end of
the write to disk, the progress indication is not as smooth as it could
be. This is a known issue and will be addressed in a future change.

This implementation may be ~0.5-1.0% slower in rare cases, depending on
network and disk speed, but is generally MUCH faster and more robust
than the its predecessor in all other cases.
This commit is contained in:
Blake Mizerany 2025-03-05 14:48:18 -08:00 • committed by GitHub
parent cae5d4d4ea
commit e2252d0fc6
No known key found for this signature in database
GPG key ID: B5690EEEBB952194
11 changed files with 370 additions and 52 deletions
Download patch file Download diff file Expand all files Collapse all files

35
server/routes.go
View file

@ -42,6 +42,12 @@ import (
"github.com/ollama/ollama/version"
)
func experimentEnabled(name string) bool {
return slices.Contains(strings.Split(os.Getenv("OLLAMA_EXPERIMENT"), ","), name)
}
var useClient2 = experimentEnabled("client2")
var mode string = gin.DebugMode
type Server struct {
@ -1173,6 +1179,7 @@ func (s *Server) GenerateRoutes(rc *ollama.Registry) (http.Handler, error) {
r.HEAD("/api/tags", s.ListHandler)
r.GET("/api/tags", s.ListHandler)
r.POST("/api/show", s.ShowHandler)
r.DELETE("/api/delete", s.DeleteHandler)
// Create
r.POST("/api/create", s.CreateHandler)
@ -1194,16 +1201,19 @@ func (s *Server) GenerateRoutes(rc *ollama.Registry) (http.Handler, error) {
r.GET("/v1/models", openai.ListMiddleware(), s.ListHandler)
r.GET("/v1/models/:model", openai.RetrieveMiddleware(), s.ShowHandler)
// wrap old with new
rs := &registry.Local{
Client: rc,
Logger: slog.Default(), // TODO(bmizerany): Take a logger, do not use slog.Default()
Fallback: r,
if rc != nil {
// wrap old with new
rs := &registry.Local{
Client: rc,
Logger: slog.Default(), // TODO(bmizerany): Take a logger, do not use slog.Default()
Fallback: r,
Prune: PruneLayers,
Prune: PruneLayers,
}
return rs, nil
}
return rs, nil
return r, nil
}
func Serve(ln net.Listener) error {
@ -1258,15 +1268,20 @@ func Serve(ln net.Listener) error {
s := &Server{addr: ln.Addr()}
rc, err := ollama.DefaultRegistry()
if err != nil {
return err
var rc *ollama.Registry
if useClient2 {
var err error
rc, err = ollama.DefaultRegistry()
if err != nil {
return err
}
}
h, err := s.GenerateRoutes(rc)
if err != nil {
return err
}
http.Handle("/", h)
ctx, done := context.WithCancel(context.Background())