chore: use original service spec instead of deriving from container, fixes diff for scale

2026-08-26 19:13:34 +00:00 · 2025-03-30 12:54:36 +10:00
parent 0975cbab66
commit 64f6a4f3d3
22 changed files with 956 additions and 405 deletions
@@ -369,6 +369,75 @@ func (c *Client) CreateServiceContainer(
 	return resp, nil
 }

+// InspectServiceContainer returns the container information and service specification that was used to create the
+// container with the given ID.
+func (c *Client) InspectServiceContainer(ctx context.Context, id string) (api.ServiceContainer, error) {
+	var resp api.ServiceContainer
+
+	grpcResp, err := c.grpcClient.InspectServiceContainer(ctx, &pb.InspectContainerRequest{Id: id})
+	if err != nil {
+		if status.Convert(err).Code() == codes.NotFound {
+			return resp, errdefs.NotFound(err)
+		}
+		return resp, err
+	}
+
+	if err = json.Unmarshal(grpcResp.Container, &resp.Container); err != nil {
+		return resp, fmt.Errorf("unmarshal container: %w", err)
+	}
+	if err = json.Unmarshal(grpcResp.ServiceSpec, &resp.ServiceSpec); err != nil {
+		return resp, fmt.Errorf("unmarshal service spec: %w", err)
+	}
+
+	return resp, nil
+}
+
+type MachineServiceContainers struct {
+	Metadata   *pb.Metadata
+	Containers []api.ServiceContainer
+}
+
+// ListServiceContainers returns all containers on requested machines that belong to the service with the given
+// name or ID. If serviceNameOrID is empty, all service containers are returned.
+func (c *Client) ListServiceContainers(
+	ctx context.Context, serviceNameOrID string, opts container.ListOptions,
+) ([]MachineServiceContainers, error) {
+	optsBytes, err := json.Marshal(opts)
+	if err != nil {
+		return nil, fmt.Errorf("marshal options: %w", err)
+	}
+
+	resp, err := c.grpcClient.ListServiceContainers(ctx, &pb.ListServiceContainersRequest{
+		ServiceId: serviceNameOrID,
+		Options:   optsBytes,
+	})
+	if err != nil {
+		return nil, err
+	}
+
+	machineContainers := make([]MachineServiceContainers, len(resp.Messages))
+	for i, msg := range resp.Messages {
+		machineContainers[i].Metadata = msg.Metadata
+		if msg.Metadata != nil && msg.Metadata.Error != "" {
+			continue
+		}
+
+		containers := make([]api.ServiceContainer, len(msg.Containers))
+		for j, sc := range msg.Containers {
+			if err = json.Unmarshal(sc.Container, &containers[j].Container); err != nil {
+				return nil, fmt.Errorf("unmarshal container: %w", err)
+			}
+			if err = json.Unmarshal(sc.ServiceSpec, &containers[j].ServiceSpec); err != nil {
+				return nil, fmt.Errorf("unmarshal service spec: %w", err)
+			}
+		}
+
+		machineContainers[i].Containers = containers
+	}
+
+	return machineContainers, nil
+}
+
 // RemoveServiceContainer stops (kills after grace period) and removes a service container with the given ID.
 // A service container is a container that has been created with CreateServiceContainer.
 func (c *Client) RemoveServiceContainer(ctx context.Context, id string, opts container.RemoveOptions) error {
@@ -2,6 +2,7 @@ package docker

 import (
 	"context"
+	"database/sql"
 	"encoding/json"
 	"errors"
 	"fmt"
@@ -455,6 +456,125 @@ func (s *Server) CreateServiceContainer(
 	return &pb.CreateContainerResponse{Response: respBytes}, nil
 }

+// InspectServiceContainer returns the container information and service specification that was used to create the
+// container with the given ID.
+func (s *Server) InspectServiceContainer(
+	ctx context.Context, req *pb.InspectContainerRequest,
+) (*pb.ServiceContainer, error) {
+	ctr, err := s.client.ContainerInspect(ctx, req.Id)
+	if err != nil {
+		if client.IsErrNotFound(err) {
+			return nil, status.Errorf(codes.NotFound, err.Error())
+		}
+		return nil, status.Errorf(codes.Internal, err.Error())
+	}
+
+	ctrBytes, err := json.Marshal(ctr)
+	if err != nil {
+		return nil, status.Errorf(codes.Internal, "marshal response: %v", err)
+	}
+
+	var specBytes []byte
+	err = s.db.QueryRowContext(ctx, `SELECT service_spec FROM containers WHERE id = $1`, ctr.ID).Scan(&specBytes)
+	if err != nil {
+		if errors.Is(err, sql.ErrNoRows) {
+			return nil, status.Errorf(codes.NotFound, "service spec not found for container: '%s'", ctr.ID)
+		}
+		return nil, status.Errorf(codes.Internal, "get service spec for container '%s' from machine database: %v",
+			ctr.ID, err)
+	}
+
+	return &pb.ServiceContainer{
+		Container:   ctrBytes,
+		ServiceSpec: specBytes,
+	}, nil
+}
+
+// ListServiceContainers returns all containers that belong to the service with the given name or ID.
+// If req.ServiceId is empty, all service containers are returned.
+func (s *Server) ListServiceContainers(
+	ctx context.Context, req *pb.ListServiceContainersRequest,
+) (*pb.ListServiceContainersResponse, error) {
+	var opts container.ListOptions
+	if len(req.Options) > 0 {
+		if err := json.Unmarshal(req.Options, &opts); err != nil {
+			return nil, status.Errorf(codes.InvalidArgument, "unmarshal options: %v", err)
+		}
+
+		// Handle filters separately because they implement custom JSON unmarshalling.
+		var raw map[string]json.RawMessage
+		if err := json.Unmarshal(req.Options, &raw); err != nil {
+			return nil, status.Errorf(codes.InvalidArgument, "unmarshal options to raw map: %v", err)
+		}
+
+		if filtersBytes, ok := raw["Filters"]; ok {
+			args, err := filters.FromJSON(string(filtersBytes))
+			if err != nil {
+				return nil, status.Errorf(codes.InvalidArgument, "unmarshal filters: %v", err)
+			}
+			opts.Filters = args
+		} else {
+			opts.Filters = filters.NewArgs()
+		}
+	}
+	// Only uncloud-managed containers that belong to some service.
+	opts.Filters.Add("label", api.LabelServiceID)
+	opts.Filters.Add("label", api.LabelManaged)
+
+	containerSummaries, err := s.client.ContainerList(ctx, opts)
+	if err != nil {
+		return nil, status.Error(codes.Internal, err.Error())
+	}
+
+	containers := make([]*pb.ServiceContainer, 0, len(containerSummaries))
+	for _, cs := range containerSummaries {
+		if req.ServiceId != "" &&
+			cs.Labels[api.LabelServiceID] != req.ServiceId && cs.Labels[api.LabelServiceName] != req.ServiceId {
+			continue
+		}
+
+		ctr, err := s.client.ContainerInspect(ctx, cs.ID)
+		if err != nil {
+			if client.IsErrNotFound(err) {
+				// The listed container may have been removed while we were inspecting other containers.
+				continue
+			}
+			return nil, status.Errorf(codes.Internal, "inspect container %s: %v", cs.ID, err)
+		}
+		ctrBytes, err := json.Marshal(ctr)
+		if err != nil {
+			return nil, status.Errorf(codes.Internal, "marshal container: %v", err)
+		}
+
+		var specBytes []byte
+		err = s.db.QueryRowContext(ctx, `SELECT service_spec FROM containers WHERE id = $1`, ctr.ID).Scan(&specBytes)
+		if err != nil {
+			if errors.Is(err, sql.ErrNoRows) {
+				// If this happens, there is a bug in the code, or someone manually removed the container from the DB,
+				// or created a managed container out of band.
+				slog.Error("Service container not found in machine database.", "id", ctr.ID)
+				// Just ignore such a container to not fail the list operation as it's not easily recoverable.
+				continue
+			}
+			return nil, status.Errorf(codes.Internal, "get service spec for container '%s' from machine database: %v",
+				ctr.ID, err)
+		}
+
+		containers = append(containers, &pb.ServiceContainer{
+			Container:   ctrBytes,
+			ServiceSpec: specBytes,
+		})
+	}
+
+	return &pb.ListServiceContainersResponse{
+		Messages: []*pb.MachineServiceContainers{
+			{
+				Containers: containers,
+			},
+		},
+	}, nil
+}
+
 // RemoveServiceContainer stops (kills after grace period) and removes a service container with the given ID.
 // The difference between this method and RemoveContainer is that it also removes the container from the machine
 // database.